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
|
|---|---|---|---|---|---|
During placenta development , a succession of complex molecular and cellular interactions between the maternal endometrium and the developing embryo ensures reproductive success . The precise mechanisms regulating this maternal-fetal crosstalk remain unknown . Our study revealed that the expression of Rac1 , a member of the Rho family of GTPases , is markedly elevated in mouse decidua on days 7 and 8 of gestation . To investigate its function in the uterus , we created mice bearing a conditional deletion of the Rac1 gene in uterine stromal cells . Ablation of Rac1 did not affect the formation of the decidua but led to fetal loss in mid gestation accompanied by extensive hemorrhage . To gain insights into the molecular pathways affected by the loss of Rac1 , we performed gene expression profiling which revealed that Rac1 signaling regulates the expression of Rab27b , another GTPase that plays a key role in targeting vesicular trafficking . Consequently , the Rac1-null decidual cells failed to secrete vascular endothelial growth factor A , which is a critical regulator of decidual angiogenesis , and insulin-like growth factor binding protein 4 , which regulates the bioavailability of insulin-like growth factors that promote proliferation and differentiation of trophoblast cell lineages in the ectoplacental cone . The lack of secretion of these key factors by Rac1-null decidua gave rise to impaired angiogenesis and dysregulated proliferation of trophoblast cells , which in turn results in overexpansion of the trophoblast giant cell lineage and disorganized placenta development . Further experiments revealed that RAC1 , the human ortholog of Rac1 , regulates the secretory activity of human endometrial stromal cells during decidualization , supporting the concept that this signaling G protein plays a central and conserved role in controlling endometrial secretory function . This study provides unique insights into the molecular mechanisms regulating endometrial secretions that mediate stromal-endothelial and stromal-trophoblast crosstalk critical for placenta development and establishment of pregnancy .
Shortly after fertilization , the uterus transitions to a receptive state that allows embryo attachment and invasion , and this process must be synchronized with embryonic development in order to ensure maximal reproductive success [1–5] . To enable this synchronization , an intricate maternal-fetal dialogue has evolved that allows the developing embryo and the uterus to be in constant communication with each other . In humans and rodents , as pregnancy progresses , the uterus undergoes a dramatic transformation to form the decidua , a stroma-derived secretory tissue that encases the growing fetus for the duration of pregnancy [3–6] . Decidual cells are responsible for producing and secreting paracrine factors that promote the formation of an extensive vascular network that supports embryo development [7–9] . Proper differentiation and migration of the trophoblast cells , critical for the formation of a functional placenta , are also influenced by as yet unknown factors secreted by the differentiating stromal cells . If any of these processes fail to proceed normally , a number of diseases of pregnancy can result , such as recurrent miscarriage , preeclampsia , and intrauterine growth restriction [10–12] . The current challenge is to understand the complex processes by which various signaling molecules emanating from the maternal decidua communicate with trophoblasts to ensure successful establishment and maintenance of pregnancy . In this study , using genetic and cell biological approaches , we demonstrate that Ras-related C3 botulinum toxin substrate 1 ( Rac1 ) , a maternal factor expressed in decidual cells , regulates the secretory pathways that mediate stromal-endothelial and stromal-trophoblast crosstalk within a narrow temporal window during placenta development . Rac1 belongs to the Rho family of GTPases and is a key signaling molecule that regulates cell proliferation , differentiation , cell-cell adhesion , and cell motility [13–16] . It controls these processes by acting as a G protein , a molecular switch that becomes active when bound to GTP or inactive when bound to GDP [13–16] . Our studies revealed that Rac1 expression is induced in decidualizing stromal cells following implantation . Conditional ablation of endometrial Rac1 led to a severe defect in fertility . Further analysis revealed that uteri lacking Rac1 are able to undergo decidualization as indicated by weight gain assay and the expression of biochemical markers of this process . However , in the absence of Rac1 , the expression of Rab27b , another G protein that plays a key role in vesicular exocytosis [17 , 18] , is markedly impaired in the decidual cells . Consistent with this finding , our studies revealed that the Rac1-null decidual cells exhibit a defect in the secretion of vascular endothelial growth factor A ( VEGFA ) and insulin-like growth factor binding protein 4 ( IGFBP4 ) . Deficiency of VEGFA in Rac1-null uteri contributed to impaired decidual angiogenesis , while the lack of action of IGFBP4 was associated with dysregulated expansion and differentiation of trophoblast cells , resulting in disorganized placenta formation and pregnancy failure . Further studies revealed that RAC1 , the human ortholog of Rac1 , regulates the secretion of VEGFA by primary human endometrial stromal cells during decidualization , highlighting its conserved role in regulating endometrial secretory function . Collectively , our study provides important insights into the molecular mechanisms that control endometrial secretions mediating stromal-endothelial and stromal-trophoblast crosstalk critical for placental development and establishment of pregnancy .
To gain insights into the molecular pathways underlying decidualization , we performed gene expression profiling to analyze alterations in uterine gene expression patterns in response to decidual stimulation . In rodents , decidualization can be induced experimentally in the absence of an implanting embryo [19 , 20] . In this protocol , non-pregnant ovariectomized mice are primed with steroid hormones and decidualization is initiated in one uterine horn by injecting oil , which mimics implanting blastocysts , while the other horn is left untreated and serves as a control . As shown previously , a robust decidual response is seen in the oil-stimulated horn by 72 h , whereas no decidualization occurs in the unstimulated horn [20] . Gene expression profiling , using RNA isolated from stimulated and unstimulated uterine horns identified many genes whose expression was significantly altered in the uterus in response to decidual stimulation ( GEO accession GSE70572 ) . Ingenuity Pathway Analysis revealed that the genes associated with cell-cell signaling , metabolism , extracellular matrix and integrin signaling , angiogenesis , and signaling by the TGFβ family and WNTs constitute the biological categories mostly affected in response to decidual stimulation . Among these factors , we focused on Rac1 because a previous in vitro study implicated that RAC1 plays a critical role during implantation in the human [21] . To confirm the results of the microarray analysis , we performed qPCR . As shown in Fig 1A , maximal Rac1 transcript levels were observed 72 h after decidual stimulation . Consistent with this finding , we observed a significant up regulation of Rac1 transcripts during early pregnancy on days 7 and 8 of normal mouse gestation ( Fig 1B ) . Rac1 , a G protein , controls downstream signaling pathways by acting as a molecular switch that becomes active when bound to GTP [13 , 15] . To determine whether the active form of Rac1 protein is present in the decidual uterus , we analyzed uterine sections on day 7 of pregnancy by performing immunofluorescence histochemistry using an antibody that specifically recognizes Rac1-GTP . We observed intense expression of active Rac1 protein in decidual cells surrounding the implanted embryo and also in the mesometrial and antimesometrial decidua ( Fig 1C ) . To investigate the function of Rac1 in the uterus , we conditionally deleted Rac1 gene in the uteri of adult mice . The conditional deletion approach was used because the global knockout of Rac1 causes embryonic lethality [22] . We crossed mice harboring the “floxed” Rac1 [23] ( Rac1f/f ) with PgrCre/+ mice to create Rac1d/d mice . This approach was previously used by several laboratories to ablate “floxed” genes selectively in cells expressing PGR ( progesterone receptor ) , including uterine cells [20 , 24–27] . We assessed the extent of deletion of Rac1 in the uteri of Rac1d/d mice by qPCR and immunofluorescence ( Fig 2 ) . Our results showed greatly reduced expression of Rac1 transcripts in the uteri on days 6 to 10 of pregnancy , indicating efficient ablation of Rac1 gene in the uteri of Rac1d/d mice ( Fig 2A ) . Consistent with the RNA profile , we observed a marked decline in the levels of active Rac1 protein in Rac1d/d uteri on day 8 of gestation ( Fig 2B ) . We further noted that the expression of other members of the Rho family of GTPases , including Rac2 , Rhoa , and Cdc42 , was unaffected in Rac1d/d uteri ( Fig 2C ) . A six-month breeding study was performed by crossing Rac1d/d or Rac1f/f females with wild-type males of proven fertility ( Table 1 ) . This breeding scheme was employed so that the implanting embryos in Rac1-deficient uteri are either intact or heterozygous at the Rac1 gene locus . At the completion of the study , we noted more than 90% reduction in the total number of pups born to Rac1d/d dams compared with the control Rac1f/f females ( Table 1 ) . The females heterozygous for the Rac1 gene delivered the same number of pups as that of Rac1f/f females . These results indicated that the severe fertility defect is attributable to the lack of Rac1 expression in PGR expressing uterine cells of Rac1d/d females . We next investigated whether the infertility of Rac1d/d females was due to an ovarian defect . Ovaries from Rac1f/f and Rac1d/d females on days 4 , 8 , and 12 of pregnancy were collected , and evaluated histologically for the presence of corpora lutea ( CL ) . As shown in S1A Fig , ovaries collected from Rac1f/f and Rac1d/d females displayed comparable histology with follicles at all stages of development and CL with normal appearance . To examine ovulation and fertilization in these mice , blastocysts were recovered from uteri of Rac1f/f and Rac1d/d mice on day 4 of pregnancy prior to implantation . No significant difference was found either in the morphology or number of the embryos recovered from Rac1f/f and Rac1d/d uteri ( S1B Fig ) . In further support of normal ovarian activity , we noted comparable levels of serum progesterone in Rac1f/f and Rac1d/d females on days 4 , 8 , 10 , and 12 of pregnancy ( S1C Fig ) . Collectively , these results indicated that the infertility of Rac1d/d females is not due to impairment in the hypothalamic-pituitary-ovarian axis or lack of fertilization but is likely due to defective uterine function . Gross examination of uterine morphology revealed apparently normal embryonic implantation sites in Rac1f/f and Rac1d/d uteri on days 6 and 8 of pregnancy ( Fig 3A ) . There was no apparent defect in uterine receptivity , embryo attachment and formation of decidual mass in pregnant Rac1d/d uteri . To further analyze the decidual response in Rac1d/d females , we performed experimentally induced decidualization . As shown in Fig 3B , both Rac1f/f and Rac1d/d uteri exhibited robust decidual responses upon stimulation . When the decidual responses were assessed by measurement of uterine wet weight gain , there was no significant difference between Rac1f/f and Rac1d/d uteri ( Fig 3C ) . Consistent with these observations , the expression of prolactin-related protein ( PRL8A2/dPRP ) and alkaline phosphatase ( ALPL ) , known biomarkers of decidualization [28–31] was comparable in the uterine sections of Rac1f/f and Rac1d/d mice on day 8 of pregnancy ( Fig 3D ) . We additionally examined the expression of a panel of factors , Pgr , Bmp2 and Gja1 ( Cx43 ) , which are known regulators of decidualization in mice [20 , 25 , 31 , 32] . Our studies showed that the expression of Pgr , Bmp2 , and Gja1 mRNAs remained unaffected by the loss of uterine Rac1 , indicating that at least certain aspects of the decidualization process progresses normally in Rac1d/d uteri ( Fig 3E ) . Although no apparent functional abnormality was detected in pregnant Rac1d/d uteri up to day 8 of gestation , we observed distinct signs of hemorrhage and embryo resorption in these uteri starting on day 10 . By day 15 of gestation , most of the embryos were resorbed in Rac1d/d uteri ( Fig 4A ) . To investigate the biological pathways affected by Rac1 deletion in the uterus , we performed gene expression profiling . Microarray analysis of decidual tissues isolated from Rac1f/f and Rac1d/d uteri on day 8 of pregnancy revealed downregulation of mRNAs corresponding to many genes among which those controlling vascular development , metabolic processes , cell differentiation , cell adhesion , and vesicular trafficking were prominent ( GEO accession GSE70446 ) . Because of the vascular defect and hemorrhage in Rac1d/d uteri , we focused on the angiogenesis-related pathways and found several factors , including Angpt2 , Nrp1 , Sphk1 , and Epas1/Hif2α , which were downregulated in Rac1d/d uteri compared to Rac1f/f uteri as early as day 8 of pregnancy ( S1 Table ) . Consistent with the microarray data , qPCR experiments validated that the expression of these factors were indeed markedly reduced , while the expression of several other angiogenic factors , such as Vegfa , Hif1α , Angpt1 , Egln1 , and their receptors , remained unaltered in Rac1d/d uteri ( Figs 4B and S2A ) . We next examined the development of vascular networks in pregnant uteri of Rac1d/d mice by employing immunofluorescence using an antibody against platelet/endothelial cell adhesion molecule 1 ( PECAM1 ) , a marker of endothelial cells . Uterine sections of the control Rac1f/f mice on day 8 of pregnancy exhibited a well-developed vascular network that spreads throughout the decidual bed surrounding the implanted embryo ( Fig 4C ) . In contrast , the PECAM1 immunostaining was markedly reduced in uterine sections of pregnant Rac1d/d mice , indicating impaired development of uterine vasculature in the absence of Rac1 signaling ( Fig 4C ) . This reduced angiogenesis was associated with considerable hemorrhagic activity in the implantation chambers of Rac1d/d mice . Staining of uterine sections of these mice with eosin-Y , which labels red blood cells ( RBCs ) , confirmed that the RBCs have extravasated from the lateral sinusoids ( S2B Fig ) . Rac1-null uteri exhibited embryo resorption on day 10 of pregnancy , but a closer histological examination of uterine sections revealed that embryos implanted in these uteri begin to show abnormalities as early as days 7–8 of gestation . Specifically , when we measured embryonic areas , we observed a significant expansion of the trophoblast cell layer in the embryos implanted in Rac1d/d uteri compared to those in Rac1f/f controls ( Fig 5A ) . Further analyses of uterine sections on days 7 and 8 of pregnancy , using antibodies against PCNA , a cell proliferation marker , confirmed enhanced proliferation of the trophoblast cells , marked by cytokeratin 8 staining , within the ectoplacental cone ( EPC ) of Rac1d/d uteri ( Fig 5B ) . We noted with interest that the timeframe of development of this embryonic phenotype in Rac1-null uteri closely overlaps with that of induction of Rac1 expression in the endometrial stromal cells during decidualization . To assess the impact of dysregulated proliferation of trophoblast cells in the EPC on placenta formation , we performed histological analyses of uterine sections of Rac1d/d and Rac1d/d mice on day 10 of gestation . As expected , the placentae of Rac1f/f mice displayed normal characteristics of one to two layers of trophoblast giant cells ( TGC ) at the maternal-fetal interface on day 10 . In contrast , we observed up to four or five layers of TGCs in Rac1d/d animals ( Fig 6A ) . The abnormal expansion of the TGCs in the placentae of Rac1d/d mice was further confirmed when we subjected uterine sections to immunofluorescence analyses using antibodies against PL1 , a TGC-specific marker [33–35] . Consistent with the results shown in Fig 6A , multiple layers of TGCs were evident at the maternal-fetal interface in Rac1d/d mice compared to one or two layers in Rac1f/f mice ( Fig 6B ) . The spatial distribution of the spongiotrophoblast cells , a subtype of TGCs , as indicated by the expression of their biomarker TPBPA [33–35] , was identical in the placentae of Rac1f/f and Rac1d/d mice ( Fig 6C ) . As pregnancy progressed to day 12 , the Rac1d/d placentae appeared to be highly disorganized , lacking properly formed layers , including the labyrinth ( Fig 6D ) . Taken together , our results indicated that decidual expression of Rac1 critically controls the proliferation and differentiation of the TGCs at the maternal-fetal interface and ensures proper placenta development and structure . We next investigated whether Rac1 signaling in the decidual cells exerts paracrine effects on the trophoblast cells during the early stages of placenta formation . Interestingly , our microarray analysis revealed that factors involved in vesicular trafficking are altered in Rac1-null decidual cells ( S2 Table ) . In particular , we observed a marked down-regulation of mRNAs corresponding to Rab27b , a member of the Rab27 subfamily of GTPases , which participate in membrane trafficking and thereby control protein secretion [17 , 18 , 36 , 37] . This subfamily consists of two closely related homologs , Rab27a and Rab27b [17] . Rab27a is expressed in a wide variety of secretory cells and participates in the exocytosis of various secretory vesicles [17] . In contrast , Rab27b expression is much more restricted and presumably tightly regulated to allow the controlled release of vesicle contents in response to appropriate physiological signals [17 , 18 , 36] . As shown in Fig 7A , we observed downregulation of Rab27b mRNA , but not Rab27a mRNA , in uterine decidual cells of Rac1d/d mice on day 8 of gestation . Consistent with this finding , we noted a marked decline in the levels of Rab27B protein in the uterine sections of Rac1d/d mice ( Fig 7B ) . Since the Rab27 proteins are known to control several steps in vesicular trafficking , including vesicle movement on tubulin cytoskeletal tracks , we considered the possibility that down-regulation of Rab27b expression in Rac1-null uteri might affect secretory activity of the decidual cells . Since Rac1d/d uteri exhibit defects in angiogenesis , we sought to determine whether the secretion of vascular-endothelial growth factor A ( VEGFA ) , a potent endothelial mitogen , was affected in Rac1-null decidual cells . To test this possibility , we employed well-established primary cultures of murine stromal cells that undergo decidualization in vitro [31] . Stromal cells isolated from Rac1f/f and Rac1d/d uteri were subjected to in vitro decidualization with estrogen and progesterone and analyzed for the expression of VEGFA . Immunocytochemical analysis revealed that in control decidual cells collected from Rac1f/f uteri , VEGFA was noticeable as diffuse staining throughout the cell , suggesting that this factor is actively trafficked through the extensive ER-Golgi-vesicular network in the cytosol prior to its secretion in the growth medium ( Fig 7C , left ) . In contrast , the stromal cells from Rac1d/d uteri displayed intense staining and retention of VEGFA within the cell , indicating impaired secretion , presumably due to a defective vesicular transport pathway . Consistent with this observation , our study revealed significantly reduced VEGFA in the conditioned media of decidual cells collected from Rac1-null uteri ( Fig 7C , right ) . Phenotypic analysis of pregnant Rac1d/d uteri revealed abnormal trophoblast expansion during placentation . A critical balance of embryonic insulin-like growth factors , IGF1 and IGF2 , and maternal insulin-like growth factor binding proteins ( IGFBPs ) , which curb the actions of these growth factors , has been shown to be important in the control of trophoblast proliferation and differentiation [38–40] . We , therefore , examined whether the secretion of IGFBPs is altered in decidual cells lacking Rac1 . Two members of the IGFBP family , IGFBP1 and IGFBP4 , were previously reported to be expressed in the mouse uterus during early pregnancy [41 , 42] . While IGFBP1 was expressed in mouse uterine epithelial cells , IGFBP4 expression was limited to stromal cells during the decidual phase of pregnancy [41 , 42] . We , therefore , determined whether secretion of IGFBP4 is affected in decidual cells lacking Rac1 . As shown in Fig 7D , there was significant accumulation of IGFBP4 in Rac1-null decidual cells , indicating a defect in the secretion of this protein . This concept received further support from the observation that the level of IGFBP4 was markedly reduced in the conditioned media of decidual cells collected from Rac1-null uteri ( Fig 7D , right ) . Taken together , our results indicated that Rac1 regulates secretion of IGFBP4 by decidual cells , which in turn would be expected to control IGF-induced trophoblast proliferation and differentiation during placentation . We next assessed whether Rac1-mediated regulation of decidual secretory pathways is conserved in the human . To test this , undifferentiated human endometrial stromal cells ( HESC ) isolated from biopsies obtained from normal fertile women in the proliferative stage of the menstrual cycle were placed in culture and subjected to decidualization in vitro in response to a hormonal cocktail containing progesterone , estrogen , and 8-bromo-cAMP as described previously [43 , 44] . As shown in Fig 8A , RAC1 transcripts are induced in HESC during in vitro decidualization . We next employed the Rac1 inhibitor Z62954982 , which specifically blocks the activation of RAC1 [45] , to investigate the role of this factor in regulating the secretory pathways in differentiating HESC . Our studies revealed that treatment of endometrial stromal cells with the RAC1-specific inhibitor did not affect the expression of transcripts corresponding to RAC1 , RAC2 , RHOA , CDC42 , or VEGFA , but led to marked suppression in the level of RAB27B transcripts ( Fig 8B ) , indicating that the regulation of RAB27B expression by RAC1 is conserved in decidua of mouse and woman . Most importantly , inactivation of RAC1 signaling and consequent down-regulation of RAB27B gene expression were associated with a strong reduction in the levels of VEGFA secreted in the conditioned media of decidualizing HESC compared to untreated HESC ( Fig 8C ) . Interestingly , while IGFBP4 levels were undetectable in the conditioned medium of decidualizing HESC , we observed significant levels of IGFBP1 in their conditioned media . The secreted IGFBP1 levels , however , remained unaffected by the presence or absence of the RAC1-inhibitor by day 8 of culture ( Fig 8D ) . These results indicated that RAC1 regulates the secretion of VEGFA , but not that of IGFBP1 , by HESC during decidualization . Collectively , our results are consistent with the hypothesis that Rac1 , acting via its downstream effector Rab27b , controls the secretory pathways that operate in decidual cells to regulate the secretion of key paracrine factors , such as VEGF , in both mouse and human endometrium .
Rac1 is a pleiotropic factor that controls a variety of cellular events and contributes to specific differentiation processes . Its GTPase activity transduces extracellular signals from seven-transmembrane protein receptors , integrins , and growth factor receptors to effector molecules that modulate multiple signaling pathways [13–15 , 46] . Activation of the mitogen-activated protein kinase pathway is a prominent mechanism that functions downstream of Rac1 in response to appropriate cellular signals [13–15 , 46 , 47] . Rac1 also promotes cell migration by regulating the formation of lamellopodia , which are sheet-like projections on the leading edge of a motile cell that propel it across a matrix [48 , 49] . Interestingly , a previous study reported that Rac1 controls stromal cell migration during invasion of human embryo into the decidua [21] . This study was limited to in vitro conditions , using cell cultures , and did not address the mechanisms via which Rac1 accomplishes this function during uterine differentiation . Our study , employing Rac1d/d mice , addressed the in vivo function of this factor in endometrial stromal cells during decidualization and embryo invasion . Surprisingly , phenotypic analysis of Rac1d/d mice did not reveal any evidence of curtailed embryo invasion during days 5–7 of pregnancy but supported a later role of Rac1 in controlling angiogenesis and trophoblast development . Our studies showed that Rac1 plays a critical role in placenta development by regulating secretory pathways in decidual cells . However , it should be emphasized that Rac1 does not regulate all types of decidual secretions . Since the loss of uterine Rac1 does not affect pregnancy until day 7 of gestation , we infer that Rac1 is unlikely to control decidual secretions during days 5–7 of gestation . Based on our results , we postulate that Rac1 acts within a critical time window that overlaps with days 8–10 of mouse pregnancy , to guide specific cellular mechanisms that regulate the decidual secretion of certain key factors , VEGFA and IGFBP4 , which influence the activity of endothelial and trophoblast cells , respectively . Mice lacking Rac1 in the decidua , therefore , present a unique model in which one can study the mechanisms by which maternal endometrial secretory pathways regulate angiogenesis and trophoblast development . A major phenotypic consequence of the loss of Rac1 signaling in the decidua is a drastic decrease in the development of the uterine vascular network that supports embryonic growth . We found that Rac1 controls angiogenesis by regulating the expression of several factors with known roles in this complex process , including neuropilin 1 ( Nrp1 ) , angiopoietin 2 ( Angpt2 ) , and sphingosine kinase 1 ( Sphk1 ) . Nrp1 , a co-receptor for VEGFA and semaphorin family members , has widespread functions in angiogenesis , axonal guidance , cell survival , migration , and invasion [50] . Angpt2 , another key regulator of angiogenesis , binds the endothelial-specific receptor tyrosine kinase 2 ( TIE2 ) to control sprouting of blood vessels . In a context-dependent manner , it can either act on TIE2-positive endothelial cells to antagonize the action of angiopoietin-1 or exert a proangiogenic effect on less mature TIE2-negative endothelial cells [51] . Sphk1 , which controls sphingolipid signaling , prevents vascular leakage during uterine angiogenesis [52] . It is conceivable that the downregulation of Sphk1 contributes to the observed vascular leakiness and hemorrhage in Rac1d/d uteri . Interestingly , Rac1 does not regulate the expression of VEGFA by decidual cells but controls its secretion . Very little is known about the mechanisms via which cells release VEGF to their surroundings , but different isoforms are known to be differentially soluble . Using green fluorescent protein-tagged VEGF , it was reported that the VEGF enters the early ER-Golgi secretory steps , but its secretion may involve trafficking mechanisms distinct from the “constitutive” secretory pathway [53 , 54] . Interestingly , a substantial fraction of VEGF-GFP is released from the cell surface by shedding , possibly as cargo contained inside extracellular vesicles [53 , 54] . This raised the possibility that Rac1 controls vesicular trafficking to regulate the secretory activity of decidual cells . Consistent with this prediction , gene expression profiling revealed that Rac1 regulates the expression of Rab27b , a member of the Rab subfamily of GTPases with restricted expression that regulates exocytic pathways of various secretory vesicles [17 , 18 , 36 , 37] . Such mechanisms allow the controlled release of dense-core granules or secretory granules , only in response to appropriate physiological signals [55] . Rab27b is known to play a critical role in regulated stimulus-induced ( vs . constitutive ) exocytosis and has been shown to control secretion of platelet dense granules and pancreatic acinar granules [18 , 36] . Since decidual cells possess secretory granules [56] , it is conceivable that Rac1-Rab27b pathway mediates their regulated exocytosis of secretory granules to control endometrial function in a stage-specific manner . Further investigation is needed to clarify the role of this signaling factor in controlling decidual cell secretions . Another major finding of this study is that decidual Rac1 regulates development of the placenta , presumably by controlling the secretion of IGFBP4 . Following implantation of the blastocyst into the uterus , trophoblastic cells in the EPC of mouse embryos must proliferate and differentiate into TGCs . It was previously shown that IGF1 promotes the proliferation of EPC cells while IGF2 induces their transformation into TGCs , which invade into the uterine tissue to gain access to the maternal blood supply . It is generally thought that autocrine secretion of IGFs by the embryo drive trophoblast proliferation and migration , whereas maternal decidua modulates their actions by secreting IGFBPs , which control IGF bioavailability . The expression of four types of IGFBPs , IGFBP1 , IGFBP2 , IGFBP3 , and IGFBP4 by human decidual cells has been reported [57] . In mice , while IGFBP-1 is detected in uterine epithelial cells , IGFBP-4 is the predominant IGFBP in stromal cells during the decidual phase of pregnancy [41 , 42] . Indeed , [125I]IGF1 ligand blot analysis of mouse uterine tissue extracts showed that only IGFBP4 was significantly increased during early pregnancy [41 , 42] . It is tempting to suggest that IGFBP4 , which is known to bind both IGF1 and IGF2 , regulates trophoblast proliferation and differentiation in the EPC by buffering the bioavailable IGFs . The Rac1d/d model , therefore , provides a plausible link between impaired secretion of a critical decidual IGFBP and the observed TGC defect in placenta development . A model depicting the role of RAC1 in endometrial angiogenesis and placental development is shown in Fig 9 . Interestingly , Nagashima et al recently reported that conditional deletion of bone morphogenetic protein receptor type 2 ( BMPR2 ) in the uterine decidua leads to abnormal vascular development , expansion of TGCs , and a deficiency of uterine natural killer cells [58] . Disruption of these pathways collectively impairs placental function and promotes fetal demise by midgestation in Bmpr2 conditional knockout mice . While there are phenotypic similarities between the Rac1d/d and Bmprd/d mouse models , it remains to be determined whether RAC1 and BMPR2 pathways converge or they function via distinct mechanisms . It is of interest that Rac1 regulation of the decidual cell secretory pathways , particularly that of VEGFA , is conserved in the mouse and the human . While the secreted IGFBP1 levels remained unaffected by the presence or absence of the RAC1-inhibitor in human endometrial stromal cells during the first 6–8 days of culture , we cannot rule out the possibility that RAC1-mediated regulation of IGFBP1 requires a longer duration of culture . Alternatively , it is possible that the human in vitro endometrial culture system does not fully capture the decidual program that occurs in vivo . Nonetheless our finding with the human endometrial stromal cultures has important clinical implications . Human placental insufficiency syndromes , including intrauterine growth restriction ( IUGR ) and preeclampsia , are characterized by abnormalities in blood vessel network formation , and trophoblast differentiation and function [59–61] . In pregnancies with IUGR , decreased expression of VEGFA and aberrant expression of angiopoietins are associated with poor placental blood vessel development [62 , 63] . Reduced expression of VEGFA has also been reported in endometrium of women with recurrent miscarriage [64] . Other studies have shown that dysregulated IGFBP1 or IGFBP4 is associated with IUGR [65 , 66] . Collectively , these results indicate that inappropriate Rac1 signaling in human decidua may result in impaired angiogenesis , vascular disruption , hemorrhage and aberrant trophoblast proliferation and differentiation , resulting in pregnancy loss . Further studies using human endometrial specimens obtained from patients with recurrent pregnancy loss may reveal whether aberrant RAC1 signaling is linked to dysregulated endometrial secretions , contributing to disruption of the maternal-embryo coordination , affecting placentation , and causing miscarriage .
Mice were maintained in the animal facility at the University of Illinois , College of Veterinary Medicine , in accordance to the institutional guidelines for the care and use of laboratory animals and in accordance with the National Institutes of Health standards for the use and care of animals . The Institutional Animal Use and Care Committee at the University of Illinois at Urbana-Champaign approved all procedures involving animal care , euthanasia , and tissue collection . Mice were housed in an animal room with temperature of 22°C and 12L:12D cycles . Food and water were provided ad libitum . Mice harboring a ‘floxed’ Rac1 gene ( Rac1tm1Glog/ tm1Glog/J , Jackson Laboratory ) were mated with mice that express Cre recombinase under the control of the progesterone receptor promoter ( Pgrtm2 ( cre ) Lyd/+ ) , termed PrgCre/+ . The Pgr-Cre mice were kindly provided by Drs . Francesco J . DeMayo and John P . Lydon of Baylor College of Medicine . The crossing of the above transgenic animals was used to produce the founding colony that produced the experimental mice containing the following genotypes: Pgr+/+ Rac1f/f ( termed Rac1f/f ) , PgrCre/+ and Rac1f/f ( termed Rac1d/d ) . This strategy has been used extensively to ablate ‘floxed’ genes in tissues expressing PGR [20 , 24–27] . Progesterone ( P4 ) , 17β-estradiol ( E2 ) , naphthol AS-MX phosphate , Fast Blue RR ( 4-benzoylamino-2 , 5-dimethoxyaniline diazonium ) , collagenase , pancreatin , dimethyl sulfoxide ( DMSO ) , 8-bromoadenosine 3' , 5'-cyclic monophosphate salt ( cAMP ) , and Trypan blue were purchased from Sigma . Hanks Balanced Salt Solution ( HBSS ) , dispase , Dulbecco’s modified Eagle medium-F12 medium HEPES , no phenol red ( DMEM/F12 ) , Penicillin-Streptomycin , Fungizone , and Phalloidin conjugated to Alexa 488 were purchased from Life Technologies . Fetal bovine serum ( FBS ) was purchased from Fisher Scientific . InSolution Rac1 Inhibitor II ( Z62954982 ) was purchased from Millipore . Fluoromount-G with DAPI was purchased from eBiosciences . Uterine sections or endometrial stromal cells were incubated with one or more of the following primary antibodies: activated RAC1 ( RAC1-GTP , 1:50 , NewEast Biosciences , 26903 ) , proliferating cell nuclear antigen ( PCNA , 1:200 , Santa Cruz Biotechnology , SC-56 ) , platelet/endothelial cell adhesion molecule 1 ( PECAM1/CD31 , 1:500 , BD Pharmigen , 557355 ) , decidual prolactin-related protein ( PRL8A2/dPRP , 1:1000 , a generous gift from Dr . Michael Soares , University of Kansas ) , vascular endothelial growth factor A ( VEGFA , 1:100 , Santa Cruz Biotechnology , SC-152 ) , cytokeratin 8 ( KRT8 , 1:50 , Developmental Studies Hybridoma Bank , TROMA-I ) , β-tubulin ( TUBB , 1:50 , Developmental Studies Hybridoma Bank , E7 ) , insulin-like growth factor-binding protein 4 ( IGFPB4 , 1:200 , Novus , NBP1-80549 ) , placental lactogen 1 ( PL1 , 1:200 , Santa Cruz Biotechnology , SC-34713 ) , Ras-related in brain 27B ( RAB27B , 1:200 , Santa Cruz Biotechnology , SC-22993 ) , and trophoblast specific protein alpha ( TPBPA , 1:200 , Abcam , ab104401 ) . The fluorescent-tagged secondary antibodies and normal donkey serum were purchased from Jackson ImmunoResearch . The following secondary antibodies were used: rhodamine or Cy3 donkey anti-rabbit , 488 donkey anti-rabbit , 488 donkey anti-mouse , 488 donkey anti-goat , and Cy3 donkey anti-rat . For immunocytochemistry , F-actin filaments were stained using phalloidin conjugated to Alexa 488 . To test fertility , Rac1f/f and Rac1d/d mice of reproductive age ( 7–8 weeks ) were paired with fertile wild-type males for six months . The total number of pups born in each litter and the number of pregnancies during this period was recorded . For experiments involving timed pregnancies , female mice were mated with adult wild-type males of known fertility . For tissue collection , all animals were euthanized by CO2 asphyxiation . Uteri and ovaries were collected at different time points during pregnancy and the tissues were immersion-fixed in 10% ( vol/vol ) neutral-buffered formalin ( NBF ) for histological evaluation or flash frozen in liquid N2 for RNA isolation or frozen sectioning . As a reference for our experiments , the identification of a copulatory plug indicated day 1 of pregnancy . Following euthanasia , blood was drawn via cardiac puncture using a 30 gauge needle and transferred into a sterile 1 . 5 mL tube . The blood samples were incubated at room temperature for 90 min to allow clot formation . After the incubation , the clot was removed with a sterile pipette tip and the samples were spun at 2000 x g for 15 min at room temperature . The serum samples were transferred into a new sterile 1 . 5 mL tube and stored at -80°C until analyzed . Serum hormones were measured by radioimmunoassay at the Ligand Core facility , University of Virginia , Charlottesville . Uterine stromal cell decidualization was experimentally induced in adult non-pregnant , hormone-primed mice as described previously [20] . Briefly , Rac1f/f and Rac1d/d female mice were ovariectomized to remove any circulating hormones . Two weeks following ovariectomy , animals were injected with 100 ng of E2 in 0 . 1 mL of corn oil subcutaneously ( sc ) every 24 h for three consecutive days . After two days of rest , sc hormones injections were given daily , containing 1 mg P4 and 10 ng E2 in 0 . 1 mL for three consecutive days . Decidualization was initiated in one horn by injecting 20 μL of oil into the lumen , while the other horn was left unstimulated . Mice were treated with additional E2 + P4 for up to 96 h post-stimulus . Mice were euthanized , uterine horns were collected and weighed . Alkaline phosphatase ( ALPL ) activity was detected following previously published protocols [67] , with modifications . Briefly , frozen uterine sections were fixed in 10% NBF for 10 min , and then washed with 1x phosphate-buffered saline ( PBS ) three times for 5 min each . The uterine sections were then incubated in the dark at 37°C for 30 min in a solution containing 0 . 5 mM naphthol AS-MX phosphate ( ALPL substrate ) and 1 . 5 mM Fast Blue RR in 0 . 1 M Tris-HCl , pH 8 . 5 . Alkaline phosphatase activity releases orthophosphate and naphthol derivatives from the ALPL substrate . The naphthol derivatives are simultaneously coupled with the diazonium salt ( Fast Blue RR ) to form a dark dye marking the site of enzyme action . The slides were rinsed in tap water to terminate the enzymatic reaction . Stained uterine sections were visualized under an Olympus BX51 microscope equipped for light imaging and connected to a Jenoptik ProgRes C14 digital camera with c-mount interface containing a 1 . 4 Megapixel CCD sensor . Mouse endometrial stromal cells ( MESC ) were isolated from uteri on day 4 pregnancy , as previously described [31] . Briefly , uteri collected from Rac1f/f and Rac1d/d female mice were cut open and digested with 5 mL/uteri with 1x HBSS solution containing 6 g/L dispase and 25 g/L pancreatin for 45min at room temperature , followed by 15 min at 37°C , with occasional mixing . After the first digestion , the supernatant , which contains epithelial cells , was removed by aspiration and the remaining pieces of tissues were washed with HBSS containing 10% ( vol/vol ) heat-inactivated fetal bovine serum ( FBS ) to stop the enzymatic digestion . The uterine fragments were washed two times with 1x HBSS to remove the serum . After the last wash , the uterine fragments were digested in 5 mL/uteri with 1x HBSS solution containing 0 . 5 g/L collagenase for 1 h at 37°C . After the second digestion , 5 mL of HBSS containing 10% FBS was added to stop the enzymatic digestion . The tubes were vortexed for 10–15 seconds until the supernatant became turbid with dispersed cells . The content of the tube is passed through an 80-μm gauze filter ( Millipore ) into a new collection tube to remove the undigested myometrial fragments . The suspension containing the endometrial cells was then spun at 430 x g for 5 min to form a pellet , and the pellet is washed with HBSS once . After the wash , the endometrial cells are spun again and resuspended in DMEM/F12 supplemented with 2% FBS , 100 units/L Penicillin , 0 . 1 g/L Streptomycin , 1 . 25 mg/L Fungizone , 10 nM E2 , and 1 μM P4 . The numbers of live cells were assessed by trypan blue staining using a hemocytometer . Mouse endometrial stromal cells were seeded in 6-well plates at an initial plating density of 5 × 105 cells . The unattached cells were removed by washing several times with HBSS , and cell culture was continued after addition of fresh culture medium . Culture medium was collected at 72 h and 96 h after the initial plating and stored at -80°C until assayed . At the end of the culture , the cells were detached from the plates , counted using Trypan Blue and a hemocytometer , and stored at -80°C for RNA extraction . The studies involving human endometrial stromal cell ( HESC ) cultures adhere to the regulations set forth for the protection of human subjects participating in clinical research and are approved by the Institutional Review Boards of Emory University , Wake Forest University , and the University of Illinois at Urbana-Champaign . Endometrial samples from the early proliferative stage of the menstrual cycle were obtained by Pipelle biopsy from regularly cycling , fertile volunteers on no hormonal medications , after providing written informed consent . HESC were maintained in a media containing DMEM/F-12 supplemented with 5% FBS , 100 units/L Penicillin , 0 . 1 g/L Streptomycin , as described previously [31 , 44] . The culture medium was changed every 48 h . Human endometrial cells were seeded in 6-well plates at an initial plating density of 2 × 105 cells , and cultured to 90% confluence . To inhibit the activation of RAC1 in HESCs , the cells were cultured in media containing 25 μM of RAC1 inhibitor , which inhibits RAC1-TIAM1 interaction [45] or DMSO control for 24 h . To induce in vitro decidualization , the cells were treated with medium containing a hormonal cocktail consisting of 10 nM E2 , 1 μM P4 , 0 . 5 mM cAMP and 25 μM of RAC1 inhibitor ( IC50 = 12 . 2 μM ) or DMSO control . The culture medium containing the inhibitor or DMSO , was changed every 48 h and the cultures were maintained for up to 8 days . Conditioned media were collected on days 6 and 8 of culture and stored at -80°C until assayed . At the end of the culture , the cells were detached from the plates , counted , and stored at -80°C for RNA extraction . Paraffin-embedded and/or frozen uterine sections were subjected to immunohistochemistry as described previously [20] . Briefly , tissues were collected and fixed in 10% NBF for 18–24 h or flash frozen in liquid nitrogen . Fixed tissues were embedded in paraffin , sectioned at 5 μm , mounted on glass slides , and incubated at 37°C overnight . Tissue sections were deparaffinized in xylene , rehydrated through a graded series of ethanol , and washed in tap water . For most of the immunostaining , antigen retrieval was performed in a pressure cooker in 10 mM sodium citrate buffer ( pH 6 . 0 ) for 20 min and then the slides were cooled to room temperature . For RAC1-GTP , specifically , antigen retrieval was performed by incubating the slides in 10 mM sodium citrate buffer ( pH 6 . 0 ) for 2 h at 80°C in a water bath . When frozen tissues were used , the sections were thawed at room temperature for 5 min and then fixed for 5 min in 10% NBF . For both paraffin-embedded and frozen tissues , washes between steps ( three times for 5 min each ) were done using 1x phosphate-buffered saline solution containing 0 . 05% Tween 20 ( PBS-T ) . Nonspecific binding was inhibited by incubating the sections with 10% normal serum for 1 h at room temperature . After the serum block , sections were incubated overnight at 4°C with the diluted antibody solution in PBS-T containing 1% normal serum . Labeling was visualized by incubation with a fluorescent-tagged secondary antibody for 1 h at room temperature . All incubations were done using a humidified chamber protected from light . Slides were mounted using a mounting solution containing DAPI . Pictures were taken using the Olympus BX51 microscope equipped for fluorescent imaging and connected to a Jenoptik ProgRes C14 digital camera with c-mount interface containing a 1 . 4 Megapixel CCD sensor . Fluorescent images were processed and merged using Adobe Photoshop Extended CS6 ( Adobe Systems ) . Primary cultures of stromal cells were subjected to immunocytochemistry as described previously [31] . Briefly , cells were fixed in 10% NBF for 10 min , and then washed with PBS . Cells were then permeabilized using PBS containing 0 . 1% Triton X for 10 min at room temperature . Nonspecific binding was inhibited by incubating the sections with 10% normal serum for 1 h at room temperature . After the serum block , the cells were incubated overnight at 4°C with the diluted antibody solution in PBS containing 1% normal serum . Labeling was visualized by incubation with a fluorescent-tagged secondary antibody for 1 h at room temperature . One drop of mounting solution containing DAPI was added to each well to stain the nucleus . Pictures were taken using the Olympus Ix70 inverted microscope adapted to a Diagnostic Instrument digital camera containing a 2 . 0 Megapixel CCD sensor . Fluorescent images were merged and processed using Adobe Photoshop Extended CS6 . Decidualization was experimentally induced in adult non-pregnant , hormone-primed mice as described previously [20] . Briefly , mice were ovariectomized to remove any circulating hormones . Two weeks following ovariectomy , animals were injected with 100 ng of E2 in 0 . 1 mL of corn oil sc every 24 h for three consecutive days . After two days of rest , sc hormones injections were given daily , containing 1 mg P4 and 10 ng E2 in 0 . 1 mL for three consecutive days . Decidualization was initiated in one horn by injecting 20 μL of oil into the lumen , while the other horn was left unstimulated . Mice were treated with additional E2 + P4 for up to 72 h post-stimulus . Mice were euthanized , uterine horns were collected and weighed . Total RNA was extracted from stimulated and unstimulated uterine horns using a standard TRIzol-based protocol . RNA integrity was verified using Agilent 2100 bioanalyser ( Agilent Technologies Inc . , Santa Clara , CA , USA ) at the Biotechnology Center of the University of Illinois , Urbana and Champaign . Each RNA sample was processed for microarray hybridization using Affymetrix GeneChip Mouse Genome 430A 2 . 0 array , which contains probes that represented approximately 14 , 000 annotated gene sequences , following the established protocol . A list of genes that had a relative fold change of >1 . 3 were further sorted by gene ontology and pathway analysis using Ingenuity Classification Software . To investigate the biological pathways affected by Rac1 deletion in the uterus , we performed gene expression profiling using uteri from Rac1f/f and Rac1d/d mice . Briefly , uterine decidual masses were isolated and freed from the embryo from day 8 pregnant Rac1f/f and Rac1d/d mice . Because Rac1d/d mice exhibit hemorrhage at this time of pregnancy , all animals were perfused with 1x HBSS containing 200 U of heparin/mL . Total RNA was extracted from the decidual masses using a standard TRIzol-based protocol . The RNA preparations were then purified with the RNeasy kit ( Qiagen ) . The purity and quality of the isolated RNA samples were assessed using an Agilent Bioanalyzer System and samples with a RNA integrity number of 10 were used . The RNA was hybridized to the Affymetrix GeneChip Mouse Genome 430A 2 . 0 array , which contains probes that represented approximately 14 , 000 annotated gene sequences . RNA quality and chip hybridization was performed by Roy J . Carver Biotechnology Center at the University of Illinois of Urbana-Champaign . Subsequently , the chips were scanned and the data were extracted using Gene-Chip Operating Software version 1 . 3 ( Affymetrix ) . A signal value for each gene below 50 was considered as background . The relative gene expression fold value was determined by the ratio of gene expression of Rac1d/d decidua to Rac1f/f decidua . A list of genes that had a relative fold change of > 1 . 3 , were further analyzed for gene ontology and functional classification using the Database for Annotation , Visualization and Integrated Discovery ( DAVID , National Institute of Allergy and Infectious Diseases , National Institutes of Health ) [68] . Total RNA was isolated from uteri , ovaries , and cells using a standard TRIzol-based protocol . The RNA concentration of each sample was determined at 260 nm using a Nanodrop ND1000 UV-Vis spectrophotometer ( Nanodrop Technologies ) . RNA samples were reverse transcribed using the High Capacity cDNA Reverse Transcription kit ( Applied Biosystems ) according to the manufacturer's instructions . Primers specific for genes of interest were developed and real-time quantitative PCR ( qPCR ) reactions were carried out using SYBR-green master mix ( Applied Biosystems ) in a 7500 Applied Biosystems Real-time PCR machine ( Applied Biosystems ) . For each sample , the mean threshold cycle ( Ct ) was calculated from Ct values obtained from three replicates . The normalized ΔCt in each sample was calculated as mean Ct of target gene subtracted by the mean Ct of the reference gene . ΔΔCt was then calculated as the difference between the ΔCt values of the control and mutant samples . The fold change of gene expression in each sample relative to a control was generated using the 2−ΔΔCt mathematical model for relative quantification of quantitative PCR [69] . The mean fold induction and SEM were calculated from at least three or more independent experiments . For HESC , the mean fold induction and SEM were calculated from at least two independent experiments . The housekeeping gene RPLP0 ( 36B4 ) , which encodes a ribosomal protein , was used as a reference gene . Reported data consists of mean fold induction ± SEM . The media of MESCs and HESCs were collected at different time points , as described above , from at least 3 wells of a 6-well plate and were stored at -80°C . Media samples were subjected to enzyme-linked immunosorbent assays ( ELISA ) and the data were analyzed according to the manufacturer's instructions . All samples were measured in duplicates and the total protein content of the culture medium was calculated . Some samples were diluted to match the dynamic range of each ELISA kit . Mean protein production was normalized to the cell counts of each sample assayed . Protein estimates were obtained from at least two independent samples for HESC and at least three independent samples for MESC . The data are reported as mean protein per cell ± S . E . M . The analytical sensitivities of each kit were: 3 pg/mL for the mouse VEGF ELISA kit ( MMV00 , R&D Systems ) , 6 . 4 pg/mL for the mouse IGFBP4 ELISA kit ( SEA055Mu , Uscn Life Science Inc . ) , 9 pg/mL for the human VEGF ELISA kit ( DDV00 , R&D Systems ) , < 5 pg/mL for the human IGFBP1 ELISA kit ( ab100539 , Abcam ) . Experimental data were collected from a minimum of four independent samples , which were subjected to the same experimental conditions . All numerical data are expressed as mean ± SEM . Statistical analysis was done using one of the following: Student’s t-test or Mann-Whitney rank sum test ( for single comparison ) , one-way analysis of variance ( ANOVA ) with a Bonferroni post-test ( for multiple comparison between samples or time points ) , or two-way ANOVA with a Bonferroni post-test ( for multiple comparison between different samples and time points ) . Analysis of equal variances was done on all numerical data to determine if a parametric or non-parametric hypothesis testing was appropriate . Data were considered statistically significant at P < 0 . 05 . All data were analyzed and plotted using GraphPad Prism 6 . 0 ( GraphPad Software ) . | During early pregnancy , a succession of molecular interactions between the uterus and the developing embryo ensures reproductive success . Although it is conceivable that signaling cues originating in the uterus impact on the developing embryo at the time of placenta establishment , the precise mechanisms regulating the maternal-fetal crosstalk remain unknown . Impaired uterine functions during early pregnancy are likely to contribute to abnormal embryo development and various diseases of pregnancy , such as recurrent miscarriage , preeclampsia , and intrauterine growth restriction . This study provides insights into the molecular mechanisms by which Rac1 , a signaling molecule expressed in the decidua , controls uterine secretions that mediate maternal-fetal communication critical for placental development and establishment of pregnancy . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2015 | Rac1 Regulates Endometrial Secretory Function to Control Placental Development |
A significant portion of bone fractures fail to heal properly , increasing healthcare costs . Advances in fracture management have slowed because translation barriers have limited generation of mechanism-based explanations for the healing process . When uncertainties are numerous , analogical modeling can be an effective strategy for developing plausible explanations of complex phenomena . We demonstrate the feasibility of engineering analogical models in software to facilitate discovery of biomimetic explanations for how fracture healing may progress . Concrete analogical models—Callus Analogs—were created using the MASON simulation toolkit . We designated a Target Region initial state within a characteristic tissue section of mouse tibia fracture at day-7 and posited a corresponding day-10 Target Region final state . The goal was to discover a coarse-grain analog mechanism that would enable the discretized initial state to transform itself into the corresponding Target Region final state , thereby providing an alternative way to study the healing process . One of nine quasi-autonomous Tissue Unit types is assigned to each grid space , which maps to an 80×80 μm region of the tissue section . All Tissue Units have an opportunity each time step to act based on individualized logic , probabilities , and information about adjacent neighbors . Action causes transition from one Tissue Unit type to another , and simulation through several thousand time steps generates a coarse-grain analog—a theory—of the healing process . We prespecified a minimum measure of success: simulated and actual Target Region states achieve ≥ 70% Similarity . We used an iterative refinement protocol to explore many combinations of Tissue Unit logic and action constraints . Workflows progressed through four stages of analog mechanisms . Similarities of 73–90% were achieved for Mechanisms 2–4 . The range of Upper-Level similarities increased to 83–94% when we allowed for uncertainty about two Tissue Unit designations . We have demonstrated how Callus Analog experiments provide domain experts with a fresh medium and tools for thinking about and understanding the fracture healing process .
Annually , there are approximately 15 million fractures in the United States , and a significant portion ( 10–15% ) fail to heal properly [1] . Both numbers and costs are predicted to increase as the population ages and as the number of osteoporosis-related fractures increases [2] . Therefore , developing intervention strategies to stimulate fracture healing is expected to positively impact health . Many of the advances made in fracture management in recent years were in mechanical stabilization and biologic bone augmentation materials such as autogenous bone graft , synthetic bone ceramics , or demineralized bone matrix [3] . The clinical impact of biological therapeutic agents , such as bone morphogenetic proteins , has fallen short of expectations for largely unknown reasons [4] . It is noteworthy that the gold standard , and most commonly used strategy for fracture nonunion treatment , autogenous bone graft , has not changed in the last 100 years [3 , 5] . Introductions of new therapeutics have slowed despite expanded research [6] . Such ineffectiveness reflects significant translation barriers . The problem is not unique to fracture-healing research; it is encountered within many research domains [7 , 8] . A translation barrier exists when mechanistic understanding of a particular medical process , such as fracture healing , is insufficient to posit a reliable , efficacious intervention strategy . A goal of the research described herein is to develop and demonstrate feasibility for a simulation-based approach , facilitating incremental improvement to a plausible mechanism-based understanding of fracture healing processes . We are not yet aspiring to utilize simulation methods to discover new mechanistic insights; knowledge is currently too sparse to support doing so . However , the approach that we employ does provide a novel means to explore and think more deeply about plausible virtual ( implemented in software ) mechanism-based fracture healing processes . Our approach is intended to be extensible to other processes that , like fracture healing , benefit from histologic analyses . We aim for our model mechanisms to follow a design such that it is straightforward to make them incrementally more biomimetic and fine-grained as new wet-lab knowledge becomes available . Before proceeding , we need a concise definition of a mechanism . In S1 Text , we provide several definitions of a mechanism which are drawn from literature sources . In support of achieving the above research goals , we are using the more detailed definition developed by Darden [9] . Paraphrasing , a biological mechanism is concrete and can be defined as a real system of entities and activities orchestrated so that it produces the phenomenon of interest , which for this work can be a feature of the fracture healing process . Thus , a model mechanism is a system of biomimetic software entities and activities organized such that , during execution , the process produces a phenomenon that is analogous to one or more features of the fracture healing process in particular ways . A model mechanism capability essential to achieving our research goal is that it facilitates hypotheses about corresponding plausible underlying features of the biological mechanism , which produces the fracture callus attribute being simulated . Fracture healing is described as comprising two phases and three stages that overlap temporally over several weeks: anabolic and catabolic phases; and inflammatory , endochondral , and coupled remodeling stages . The dominant cell types and subprocesses [10] change as healing progresses . Recent analyses of transcriptomes present during fracture healing have shown that most of the genes and signaling pathways that are involved in skeletal development in embryos are also expressed in cells of the fracture callus [11] . Consequently , some pathway components have become the focus of empirical research efforts to develop therapeutic interventions [10] , despite the fact that there is no model of explanation—even at a coarse-grain—for stages in the fracture healing process . Core phenomena of embryogenesis and some types of tissue regeneration include the evolving small- and large-scale patterns that are readily apparent in recorded images . There has been considerable progress in developing mechanism-oriented explanations for those phenomena [12] . However , stained tissue sections of mouse tibia fractures obtained at intervals of several days lack the hallmarks of orderly , organized evolving phenomena exhibited by embryogenesis . The strikingly less organized callus tissue obscures the ongoing order of the various subprocesses and their mechanisms . Part of the problem traces to limitations of experimentation . Healing of mouse tibia fractures typically spans four-to-five weeks . A major complication is that , within the same experiment , no two fractures are the same . Although the healing phenomenon is the same , the unfolding healing subprocesses within each callus are unique . Large observational gaps coupled with the necessary limitations of standard histological techniques means that informative subprocesses or phenomena may be missed . It is also plausible that informative phenomena—patterns and features—are being observed and recorded , but are not yet recognized as such . Analogous circumstances have existed in non-biological domains , and significant progress has been achieved using computational and grid-based simulation methods to provide plausible model representations of the missing processes and phenomena . For example , looking for improved insight into processes occurring at the interface of ecology and geomorphology , Fonstad opined , “we have thousands of such images , but no theories in geomorphology nor ecology can fully explain the patterns in any of them” [13] . The fact that callus mechanisms have been successfully healing bone fractures for more than 150 million years [14] implies the existence of a well-orchestrated , robust process . Similarities of callus and embryonic transcriptomes support that inference [2] . If we accept the premise that fracture healing is a well-orchestrated , robust process , then we need to answer this question: how can we begin developing a theory about the healing process—even if initially coarse and somewhat abstract—so that we can begin theorizing about its orchestration ? A clearly described phenomenon is a precondition for developing a theory intended to explain that phenomenon ( S1 Text ) . However , we do not yet have a clear temporal description of the fracture healing process , or even for portions of the process . We do , however , have detailed descriptions of features of the process at different stages . With current technology , it is not feasible to measure a callus continuously . Likewise , it is infeasible to track the changing variety of local structures and cell types . Must we plead for more data , and then come back to the problem in another decade or two ? Absent a plausible explanation and theory to test , more data may not be the answer . In discussing comparable issues at the ecology-geomorphology interface , Fonstad observed that , “both of these disciplines are data-rich … it is immediately apparent that both of these disciplines are far more theory-poor” [13] . Fracture-healing research is handicapped because it is relatively data-poor and theory-poor . So , although we can draw inspiration from the explanatory , pattern-oriented simulation methods used by Fonstad and others , their models and those pattern-oriented techniques are not yet applicable in advancing fracture-healing research . Given the growing interest in increasing the clinical relevance of modeling and simulation research , it is not surprising that the number of such reports in which authors utilize histology images to support face validation and/or guide calibrations is also increasing . The following are three recent examples . Marino et al . [15] utilized their model of lung granuloma formation to compare in silico granulomas to those of the nonhuman primate Macaca fascicularis . Gardiner et al . [16] utilized an agent-based particle system at various granularities to simulate mechanical behaviors of cells and tissues . Simulations using selected parameterizations bore a close resemblance to histological observations of an epithelial layer , cell clusters , and single cells . Ziraldo et al . described an agent-based model of ischemia/reperfusion-induced inflammation coupled with pressure ulcer formation and progression in humans with a spinal cord injuries [17] . Serial photographic images spanning several clinical stages were used to calibrate progression and healing of virtual pressure ulcers . Virtual pressure ulcers were interrogated to explore how and when a irritation might resolve or become chronic . The prospect of pulling together a start-to-finish tissue-level mechanism-oriented description of a fracture healing process , even one that is initially coarse-grain , seems distant . Why ? It is a consequence of four interrelated obstacles arising from fracture-healing research using rodent models . Given those obstacles , current knowledge and methods are insufficient to describe , much less begin building a conventional molecular and cellular biology-based model of fracture healing . The most pressing current need is to develop strategies and methods to circumvent and eventually overcome each of the above four obstacles . We conjectured that the software-based model mechanism methods , which we have used successfully in other contexts ( e . g . , see [22–27] ) , could provide the foundation for such strategies , even though , for those earlier applications , considerably more mechanism related fine-grain knowledge was available . Briefly stated , the cited software-based model mechanism approach begins with a target phenomenon . We build an extant ( actually existing , observable ) , working mechanism in software that is parsimonious and , based on similarity criteria , exhibits essentially the same phenomenon . Doing so requires making no assumptions about the biology . However , even when the mechanism is kept coarse-grain , the space of possible software mechanisms capable of generating essentially the same phenomenon can be huge . So , biologically inspired requirements and constraints along with mechanism granularity limits are imposed incrementally to shrink and constrain possible model mechanism space . That process shrinks a large set of possible coarse-grain mechanisms into a much smaller set of plausible , incrementally more likely and increasingly biomimetic , model mechanisms . For fracture healing , we envision simulations generating plausible scenarios for how discretized features of a callus tissue section on one day might transform progressively into the tissue section features—target features—observed several days later . Wet-lab experiments can target differences in two model mechanisms , where the resulting new evidence is expected to support one mechanism and falsify the other ( as in [28] ) , further shrinking plausible mechanism space . At that stage , the surviving software mechanism can stand as a coarse-grain theory for how a portion of mouse tibia fracture healing occurs . Eroding the four obstacles in meaningful ways requires coupling the preceding methods with an important new capability: use of image interpolation strategies to build plausible sequential image models of the same fracture at different stages of the healing process . A prerequisite for an interpolation strategy is having and aligning discretized coarse-grain models of tissue section images of tibia fractures from different mice at different times . We report results of a focused demonstration that meets the above requirements . We present results of workflows that support the feasibility of the approach , while also bringing its weaknesses into focus . For this demonstration , we limited attention to the critical interval from day-7 to day-10 during healing of a mouse tibia fracture and focused on discretized models of specific tissue sections on both days . From the latter , we obtained the initial state and final target state for our simulations . Biomimetic software mechanisms involving actions of quasi-autonomous tissue units spanning , typically , 5 , 000–6 , 000 time steps are responsible for simulated healing . Similarities ( defined in Methods ) between simulated and referent final states ranged from > 73% to > 93% , depending on the nature and stringency of the Similarity criterion . Despite the narrow focus , it is clear that a major benefit of the approach demonstrates that simulation experiments can enable discovering , challenging , and improving theories of healing subprocesses . Because our approach and methods are unconventional , somewhat new , and still evolving , we present that information next under Methods to provide the context needed to present and discuss results . There are weaknesses and limitations associated with every aspect of our approach . Some are identified in Methods , and others are addressed under Discussion . We undertook this demonstration with the expectation that the more successful methods could be repurposed to begin lowering similar barriers faced within some domains of disease progression research .
We begin with a synopsis of our approach from a workflow perspective , as diagrammed in Fig 1 . We then provide details on methods used during each of the six stages . In several places , we also provide essential background information that influenced decisions for a particular stage . Words , such as tissue , mechanism , healing , and process , are used in discussing actual mouse tibia fracture healing and corresponding simulations . To reduce confusion , we capitalize those words hereafter when discussing Callus Analogs . We focus on the day-7 to day-10 interval of mouse tibia fracture healing because histomorphological evidence indicates that the relative contributions of chondrogenesis and osteogenesis may undergo important changes during that interval . The goal is to develop a concrete , quantitative ( and thus challengeable ) but partially coarse-grain theory that may explain how characteristic tissue level features on day-7 are being transformed into corresponding features observed on day-10 . The discovery effort would be greatly simplified if we could obtain collocated day-7 and day-10 tissue sections from the same callus ( mouse 1 ) , but that is infeasible . Instead , we used an evidence-based illustration ( created by coauthor M . M . ) of an envisioned tissue section of the mouse 1 callus on day-10 at the same callus location as the day-7 tissue section . Three domain experts ( see Acknowledgments ) judged it plausible and acceptable . Hereafter , we refer to the illustration as the day-10i tissue section . A square grid was used to discretize the day-7 and day-10i images . The area of tissue at each grid location was labeled as one of nine tissue types , based on staining and preponderance of cell types within that space . The result ( stage 2 ) was a discretized coarse-grain model of the day-7 and day-10i tissue sections . Because we are at the beginning of this explanatory discovery process , we needed to select a target region on which to focus ( discussed further under Target Region ) . From a simulation perspective , the target region has an initial state , which maps to the day-7 tissue section , and a corresponding final state , which maps to day-10i tissue section . During stage 4 we used the MASON simulation toolkit [29] to create a 2D 25×25 Target Region initial state , in which objects representing tissue units are assigned to each grid space . We start with a 2D Target Region to limit uncertainty in tissue type identification and to adhere to our parsimony guideline . Stage 5 efforts focused on answering the following question: how do we enable the Target Region initial state to transform itself so that the arrangement of tissue types mimics the Target Region final state ? The steps followed to answer that question involved iterative refinements ( discussed below ) and had two objectives . 1 ) Explore logic to be used by simulated tissue units that enable them to successfully transition into biomimetic final states . 2 ) In doing so , keep the logic simple and avoid process features that may appear non-biomimetic . Once we had evidence that reasonably biomimetic final states were achievable , we shifted attention to improving the simulated healing process sufficiently to achieve the following quantitative target Similarity value ( stage 6 ) : compositional and organizational similarity between simulated Target Region and day-10i final state is ≥ 70% . So doing would support the feasibility of achieving the long-term Fig 1 goals . A simulation that uses concrete objects ( simulated tissue units ) to generate a process that is analogous to callus healing in several ways is a software analog of the healing process . We call the parameterized software a Callus Subregion Analog . Hereafter , for convenience , we refer to the software as Callus Analog and , in some places , simply Analog . Histologic slides of sagittal sections through mouse tibia calluses at various stages of healing were available from a previous study . The sections were stained using Hall-Brunt Quadruple to highlight tissue , bone , and cartilage . Shown in Fig 2 are the tissue sections from mouse 1 on day-7 and from mouse 2 on day-10 . Coauthor R . M . selected them because they have similar fracture features and exhibit all characteristic callus features . The following nine distinct microscopic tissue types are common to all normal mouse tibia calluses , beginning before day-7 and extending beyond day-10 . We assigned a different color to each tissue type , which was used to colorize a discretized version of Fig 2A . A microscopic area of callus containing about 20 or more cells can be distinguished as being either new marrow ( 4 ) , new bone ( 5 ) , hypertrophic cartilage ( 6 ) , mature cartilage ( 7 ) , or young cartilage ( 8 ) based on the characteristic heterogeneous mix of cell types , the dominant cell type , and extracellular matrix . As healing progresses the mix of cell types within a microscopic area changes . Some areas may undergo multiple tissue type transitions . A working hypothesis is that each of the microscopic tissue types is engaged in somewhat different activities , which are integral to the overall healing process . The first stage 2 task was to select a square grid mesh size and overlay it on Fig 2A . Choice of mesh size was somewhat arbitrary . If it is too fine , there are fewer cells within the microscopic area and so the uncertainty in specifying the dominant cell type increases . If too coarse , the fraction of microscopic areas containing clearly distinguishable tissue types 4–8 decreases , rendering a single tissue assignment inadequate ( and actions of the analog counterpart would likely require unique logic ) . A guideline for selecting grid size was that the cellular heterogeneity observed within the larger local callus area be reasonably preserved in the discretized image . For example , for a macroscopic region characterized by a heterogeneous mix of predominately ~ 60% new marrow ( gray ) and ~ 40% osteoblasts ( burgundy; new bone ) , the discretized image counterpart should be a mix of ~ 60% gray and ~ 40% burgundy tissue units . We selected a mesh size that corresponds to an 80×80 μm area in Fig 2A , which typically contained roughly 40 cells , and overlaid that grid on the day-7 and day-10i tissue sections . We then designated each microscopic area as being one of nine concrete , quasi-autonomous , Tissue Unit types , where the behavior of each Tissue Unit type was controlled by a software agent . Fig 3 contains the resulting discretized , colorized images . Although physically correct image interpolation ( e . g . , between day-7 and day-10i ) is infeasible , sophisticated image interpolation methods , as demonstrated by Stich et al . [30] , are available to create high-quality , convincing model images that represent unobserved transitions between recorded images of the same object . The criterion for an acceptable interpolation used by Stich et al . , is qualitative: the interpolated images are perceived as visually correct by human observers . During stage 1 , we faced the more daunting problem identified in Fig 4: we needed an image that plausibly anticipates the appearance of the mouse 1 fracture if it had been sectioned on day-10 rather than day-7 . Starting with the features evident in Fig 2A , and drawing on the tissue features in Fig 2B , coauthor M . M . created an illustration of the envisioned mouse 1 , day-10i section . It was judged plausible and acceptable by coauthor R . M . and , separately , by three independent domain experts ( see Acknowledgments ) , thus concluding stage 1 . Clearly , a different medical illustrator , one knowledgeable about callus progression , would create a somewhat different illustration . However , we suggest that variability introduced by such illustrations will not add measurably to the considerable variability and uncertainties already present , as illustrated by Fig 4 . To demonstrate feasibility , we needed to designate a Target Region , but first , we needed to select a portion of Fig 2A in which to locate the Target Region . For the latter , we selected the yellow-boxed area in Fig 2A . It is bordered on one side by bone and marrow cavity , which means that transitions in that area will be focused rightward , rather than occurring in two or more directions . Because that area , and the corresponding region in Fig 2B , exhibit similarities , we conjectured that the variety of feature changes occurring during transition from day-7 to day-10i might be representative of key healing features occurring elsewhere in the callus during that 4-day interval . There is no indication that unique healing features may be occurring within this area but not elsewhere during that 4-day interval . Specifying the size of the target region is subject to opposing constraints . If the region is too large , with a large variety of tissue transition types , we run the risk that the process of discovering plausible and parsimonious logic to direct transitions will become unwieldy , possibly even problematic . If the region is too small , the variety of transition types may be too few to enable adequately simulating Target Region final state . We selected the 25×25 grid region designated by the white box in Fig 3A . Fig 3B shows the corresponding Target Region final state . Limiting attention to just one target region can be viewed as a weakness . On the contrary , it is an essential part of a recognized , long-term mechanism-discovery strategy that can build on methodological lessons learned while using the Iterative Refinement Protocol in other contexts [23 , 25 , 27 , 28 , 31] . That strategy employes variations of the forward/backward chaining ( described in S1 Text ) and requires selecting a Target Region ( stage 3 , Fig 1 ) . After we achieve stage 6 for the day-10i Target Region ( described below ) , we envision expanding the temporal reach of Callus Analog Mechanisms along the dotted line illustrated in Fig 4 to include an earlier stage within that same Target Region , such as day-4i , and a later stage , such as day-14i , and doing so all while continuing to simulate the original day-10i Target Region . Those objectives are illustrated by two unshaded bars labeled a and b in Fig 4 . Further , the histological evidence suggests that , on the same day , different subregions within a callus can be at somewhat different stages of repair and may progress at different rates . Given that , a parsimonious strategy is to select separated target regions within the same callus and develop simulations for each in sequence . They could be treated as independent modules . Future work based on simulations of independent target regions will help bring regional issues into focus prior to engineering their merger . The process of merging initially independent modules into a unified model of a tissue healing process would occur further downstream . Given that this work strives to establish the feasibility of the Fig 1 approach , it is efficient to focus first on one Target Region . Simulation requirements—and thus software requirements—flow directly from desired use cases [32] . In the Introduction , we stated that a primary use case is exploratory simulations capable of the following: aiding image interpolation and providing plausible explanations for how callus features are progressively transformed , all while shrinking the space of possible explanatory transformation scenarios . The last two requirements involve generation of plausible mechanism-oriented explanations , illustrated in Fig 1 . To realize use cases , we employ the virtual experiment approach described by Kirschner et al . [33] , along with enhancements drawn from Smith et al . [28] and Petersen et al . [31] . In doing so , the methods employed must meet the following three requirements , which are based on broader sets of requirements discussed by Hunt et al . [32] . To achieve requirement 2 , Callus Analogs are written in Java , utilizing the MASON multi-agent simulation toolkit [29] . The data presented herein along with Callus Analog code are available [36] . We customized the established Iterative Refinement ( IR ) Protocol [22 , 27 , 28 , 31 , 32] to meet the challenges evident in Fig 4 . Given a software Mechanism that may explain a specified attribute and a virtual experiment design , the goal of an IR Protocol cycle is to test this hypothesis: upon execution , simulation features will mimic the target attribute within a prespecified tolerance . A concrete software mechanism can be falsified—shown to be inadequate—when , during the course of many Monte Carlo trials , it too often fails quantitatively to achieve its objective and/or exhibits non-biomimetic features . It is from encountering and overcoming such failures that explanatory insight improves . Each falsification improves credibility incrementally and shrinks plausible Mechanism space . Our customized IR Protocol follows: Well-organized processes are responsible for the callus remodeling that occurs between day-7 and day-10 . Our operating hypothesis is that information available in day-7 and day-10 tissue section images can be used to draw simplified inferences about unobserved transitions that occur during intervening days , analogous to the approach used by Stich et al . in simulating unobserved transitions between recorded images of the same object [30] . A Callus Analog Mechanism is a system of biomimetic software entities and activities organized such that , during execution , it produces a representation that is measurably similar to the day-10i Target Region . Feature changes within the Target Region explain how the Phenomenon is generated . Stained tissue sections provide snapshots of the healing process . To be explanatory , a biological mechanism will exhibit the fourteen features identified in S1 Text . Because Callus Analog Mechanisms exhibit those same features ( also identified in S1 Text ) , the two processes may be analogous . Simulations are discrete time; time advances in steps . Fig 5 shows the Target Region initial and final state . The Process responsible for transitioning from initial to final state is the top-level Mechanism . Changes within local subregions from one time step to the next are lower level Phenomena . The lower level Mechanisms responsible for those changes are characterized by individual TU changes , which are controlled by the logic that governs TU agent actions during each time step , discussed below . During each time step , each TU agent , selected randomly , has one opportunity to update and act , based on changes that have occurred within its Moore neighborhood since it last updated . An action may change a TU type or one of its Moore neighbors . Coauthor R . M . identified the following as allowed but not required biomimetic transitions . At each time step , the current simulated Target Region is compared to the Target Region final state and percent Similarity is calculated as follows: %Similarity=100 ( ∑ ( i=2−8 ) ( CSi/CFi ) n ( CSi/428 ) ) t , where t designates the time step , and i specifies the TU type , 2–8 . CSi is the count of TU type i in the simulated Target Region; CFi is the count of TU type i in Target Region final state; 428 is the number of active TUs in Target Region; n = 1 if CFi > CSi , and n = –1 if CSi > CFi . A case can be made that , if there is strong similarity between gray and burgundy TUs in the simulated and actual Target Region final states , then the similarity score should not be penalized because there are too many simulated gray TUs and too few simulated burgundy TUs , or vice versa . New marrow ( gray ) and new bone ( burgundy ) are always formed together . Thus , in some cases , the decision to designate an 80×80 μm area of stained tissue section as either gray or burgundy can be arbitrary; two experts may make different assignments . There are several ways to address that issue but they involve adding at least one new TU type . Given our strong parsimony guideline and the fact that we are at a very early stage in developing the Fig 1 approach , we elected to also calculate a Similarity value when gray are burgundy treated as the same during the calculation . The resulting value is designated Upper-Limit Similarity , simply UL-Similarity hereafter . A more realistic value may be between Similarity and UL-Similarity . When working to discover plausible model mechanisms , there is a risk that the modeler , subconsciously or otherwise , will favor Mechanism features and logic that ensure that outcomes of generated behaviors are as the modeler thinks that they should be . We strove to eliminate that risk by adhering to the guidelines in steps 4 and 6 of the IR Protocol . In developing model mechanisms , we did not aim to include established biological features . Instead , we worked to develop model mechanisms that did not contradict known biology . Along the same lines , our model mechanisms were developed not to specifically describe characteristics of the fracture healing process , but instead to allow for a new , coarse-grained manner in which to think about the process . Additional information about the disadvantages of absolute grounding can be found in [40] . By adhering to a strong parsimony guideline ( IR Protocol step 2 ) , we avoided adding unnecessary details; doing so enabled us to avoid inscription error . More details on overfitting and analog-to-referent mappings can be found in the Discussion section of Kim et al . [26] . In the subsections that follow , we describe four workflow stages , designated Mechanisms 1–4 . What further improvements in similarity values , as calculated above , might reasonably be achieved ? We answered that question with Mechanism 4 internal control calculations that draw on the fact that the closest Similarities that can be achieved for a simulated day-10i Target Region will be those achieved by independent executions of the Mechanism that generated it . The analog Healing Process from day-7 initial state to a simulated day-10 final state is unique for each Monte Carlo execution of Mechanism 4 . We selected one Mechanism 4 Monte Carlo execution from 25 and recorded its Target Region configuration at the time step for which simulated Target Region final state maximum Similarity was achieved . We designated it to be the internal control simulated day-10i target state . We then measured maximum Similarity of each of the other 24 Monte Carlo Healing Processes to that simulated day-10i target state . Data are available and labeled as “Internal Control” in S1 Dataset . Callus Analogs are a form of data , using both the implicit schema of MASON/Java and the explicit schema of configurations . Analog and configuration data are maintained , archived , and released using the Subversion version control tool in two repositories: one public for collaboration and one private ( Assembla ) for rapid and prototyping development with project partners [36 , 41] . Input-output ( I/O ) data is handled separately . Smaller data sets ( tissue data ) are stored in CSV format . The entire Callus Analog toolchain is open-source , thereby enabling repeatability . Similarly , all generated and released data from the project is licensed and available as open data . Callus Analogs are built and maintained for a cloud environment ( e . g . , Google Compute Engine ) to ensure platform and infrastructure repeatability across future experiments , project team members , partners , and the wider community . The protocols , procedures , and methods that we employ to ensure that results of Callus Analog experiments are reproducible and to establish credibility that they are scientifically useful are described in detail in [28] . Except as noted , we followed those best practices during the workflows described above . They include 1 ) quality assurance and control protocols , 2 ) face validation , 3 ) verification procedures for model mechanisms , 4 ) repeatability , 5 ) methods for generating narrowly focused predictions , and 6 ) Callus Analog validation methods used during IR Protocol cycles . However , it is too early for systematic sensitivity analyses or efforts to quantify uncertainties associated with particular Mechanism 4 configurations . That is because Callus Analog is still at a very early stage; the focus is on acquiring new insights . Mechanism life cycles are expected to remain relatively short . As soon as we target additional attributes , it is likely that Mechanism 4 will be falsified ( because it cannot achieve those new targets ) . Thus , it will be necessary to alter Mechanism 4 during additional cycles through the IR Protocol . We acquire evidence that we are adhering to our strong parsimony guideline at IR Protocol step 2 in part through documentation of sources of uncertainty and focused assessments of simulated final state sensitivities to modest changes in the logic followed by each TU .
Each execution of our Mechanisms is a unique , simulated healing process , which may ( or not ) mimic an interval of actual fracture healing . Because the focus is on similarity to the day-10i Target Region , we refer to the simulated state having maximum Similarity value for a given Monte Carlo execution as the simulated day-10i Target Region final state . However , we currently have no data to guide mapping time steps to wet-lab clock time . Summarized maximum Similarity results for Mechanisms 2–4 are included in Table 1 and the complete results are included in S1 Dataset . Examples of simulated Target Region final states for Mechanisms 2–4 are provided in Fig 9 . S1 Video is an example of the complete simulated healing process for Mechanisms 2 . It includes the final state having the largest Similarity value . Mechanism 2 improved upon Mechanism 1 , generated reasonable simulated states , but failed to meet the biomimesis requirement . A feature of a simulated healing process that has no known real counterpart is designated non-biomimetic . Fig 9A shows the simulated Target Region having the largest maximum Similarity value from a set of 25 executions . Although maximum Similarity values over 70% were achieved for Mechanism 2 , at least two non-biomimetic features were observed . 1 ) In simulated final states , teal TUs were absent from the southeast region of Target Region and that reduced Similarity values significantly . The small islands of teal and blue TUs within a gray/burgundy region in Fig 9A may also be non-biomimetic features . 2 ) Temporal profiles of Similarity values reached a plateau prior to time step 5000 , which persisted beyond time step 20 , 000 . Consequently , the time step at which maximum Similarity occurred appeared somewhat random . Fig 9B shows the simulated Target Region having the largest Mechanisms 3 maximum Similarity value from a set of 25 executions . S2 Video includes Fig 9B . Table 1 data shows that Mechanism 3 improved on Mechanism 2 . However , the absence of blue TUs adjacent to teal in the lower right limited maximum Similarity values and may be a non-biomimetic feature . Fig 9C and 9D , which are included in S3 and S4 Videos , are examples of simulated Target Regions having the largest Mechanisms 4 maximum Similarity value from the set of 25 executions summarized in Table 1 . No non-biomimetic features were observed . Based on simple qualitative visual comparisons , we judged the similarity of Fig 9E–9H to the day-10i Target Region to be comparable to that of Fig 9C and 9D , even though their Target Regions had smaller maximum Similarity values . That observation indicates that , moving forward , improved measurements of Similarity will be needed . S5 and S6 Videos includes Fig 9E and 9F . Each Mechanism 4 is a unique , simulated healing process , which is intended to mimic the interindividual variability of actual fracture healing . To observe and measure that uniqueness , single Mechanism 4 executions were selected from those used to provide the summary results for Mechanism 4 in Table 1 . A Similarity value was calculated and plotted each time step in Fig 10 . Looking for such features and for wet-lab evidence that pushes the decision either way should be part of future research Callus Analog research . We saved the images corresponding to the maximum Similarity for both sets of 25 Mechanism 4 executions summarized in Table 1 . Similarity value is just one way to judge good overall biomimicry . Assessments of simulated final state sensitivities to modest changes in TU logic help identify sources of uncertainty . They also provide evidence for how tightly we are adhering to our strong parsimony guideline . For example , changing action and event probabilities in Mechanism 4 by ± 10–15% for a particular TU produces changes in Similarity value for simulated final state and temporal profiles that are well within the range produced by 25 Monte Carlo executions; the behavior space of Mechanism 4 is not significantly altered . The following is a specific illustration . We changed the Moore neighborhood probability values for blue ( randomly chosen ) in Fig 8A from [0 . 2 north/south and 0 . 6 east] to [0 . 165 north/south and 0 . 67 east] and repeated the 25 executions tabulated Table 1 using the same seeds . The average maximum UL-Similarity was 93 . 6% vs . 93 . 3% in Table 1; and for Similarity it was 77 . 8% vs . 77 . 6% in Table 1 . For mean time step at which those value occurred , the new ( vs . Table 1 ) time step was 5317 ( vs . 5274 ) for UL-Similarity and 5299 ( vs . 5244 ) for Similarity . However , changing how and when invasion of TUs from the north is triggered is an example of a change that can have a more significant influence: for such changes , the behavior space of Mechanism 4 can be significantly altered . Data labeled “Blue Probability 2” in S1 Dataset reflect this experiment .
Each Callus Analog execution provides a record of an analog Healing Process , which is the top-level analog phenomenon . Executions generate the succession of changes by which earlier states of the Target Region gradually become a simulated Target Region final state . Each video explains how the initial state is transformed into a final state that is measurably similar to Target Region final state . It is too early to claim that strong analogies exist between features of simulated and actual fracture healing for Mechanism 4 . Nevertheless , we can hypothesize that , at comparable levels of granularity , the simulated healing processes seen in the S3–S6 Videos have actual fracture healing counterparts . Taken together , each video is a low-resolution ( coarse-grain ) model of explanation—a theory—that maps to a 4-day portion—day-7 to day-10—of the tibia fracture healing process in a mouse . There are currently no comparably detailed competing theories of fracture healing . Because Callus Analog mechanisms are concrete , they are easy targets for scientific challenge , and it is through that use that we anticipate Callus Analogs will provide scientific value moving forward . The changes occurring within the Target Region during Mechanism 4 executions are intermediate level phenomena; they have histomorphometric counterparts in callus tissue sections . Two examples are the eastward progression of teal TUs replacing blue TUs and the eastward expansion of the mixture of gray and burgundy TUs . The mechanisms responsible for those intermediate level phenomena are mediated by the individual activities of the participating TUs . The logic dictating TU actions at each time step , as illustrated in Fig 7C , provides Mechanism orchestration . A change in TU type at a particular grid location is the lowest level ( finest-grain ) model mechanism phenomenon . Mechanisms 1–4 were developed sequentially . Derived results were most useful when a particular configuration ( logic , utilization of neighborhood information , probability values , etc . ) failed to meet expectations . Some failures were marked by poor maximum Similarity values . Others were marked by a feature or features within Target Region that was unexpected or judged non-biomimetic . In such cases , we would hypothesize a plausible explanation for the problem and a possible solution , and then conduct experiments to challenge that hypothesis and solution . When successful , we achieved an incremental Analog improvement . Failures provided new knowledge by enabling us to marginally shrink the space of plausible Callus Analog healing processes . Mechanism 1 was unsuccessful because it was flawed in several ways . Nevertheless , observations made during IR Protocol cycles stimulated ideas for other logic that might be explored , including the ideas that drove development of Mechanism 2 and , later , Mechanism 3 . From observations made during explorations leading to Mechanisms 1 and 2 , we inferred that , in order to make simulated healing processes more biomimetic , it would be necessary to include two features . 1 ) Allow for multiple TU changes at any grid location during simulated healing . 2 ) Have sustained directional influences , spanning many TUs , guiding or constraining the direction and type of TU transitions . The latter may map to the combined net effects of multiple factors such as angiogenic impairment [44] , relative abundance and activity of immune cells [45] , signaling influencing osteogenic and chondrogenic transcription networks [46] , O2 gradients [47] , and mechanical influences [48] . Callus Analog has achieved its current objectives without needing to bring any of those influences into focus , consistent with our strong parsimony guideline . As the list of targeted phenomena expands , it will become necessary to make model mechanisms finer-grain . It is during such refinements that a newly added Callus Analog feature may map to one or more of those influences . The logic used by Mechanisms 3 and 4 limits the direction in which a TU can affect the transition of a neighbor , and it imposes preconditions on number and type of Moore neighbors that must be present before a TU transition can occur . A consequence of those constraints was the emergence of apparent cohesion of TUs within three areas that are clearly evident in Fig 9E–9H: one area dominated by blue TUs , another dominated by teal TUs , and a third dominated by gray and burgundy TUs . That apparent cohesion is clearly evident during S2–S6 Videos . From the simulation engineering perspective , Callus Analog could be simplified if those three areas were represented as large , quasi-autonomous , sub-Callus organized units , within which TUs are simply parts under the control of each organized unit . However , there is , as yet , no direct wet-lab evidence that would support or require such simplification . Interfaces between those areas map to well-documented transition zones ( e . g . , see [39] ) . Moving forward , features of transition zones will be added to an expanding list of targeted phenomena to further shrink the space of plausible explanatory model mechanisms . Special attention was given to understanding why , during an IR Protocol cycle , a model mechanism failed . As Petersen states , "having that information is essential to the scientific process because it is falsification that provides new knowledge: specifically , the current ( falsified ) mechanisms are flawed—they are not a good analogy of the referent biological mechanisms" [31] . Building upon and revising flawed hypotheses offers a new perspective and new way of thinking about plausible networked callus healing processes , and that alternative way of thinking may well become the primary value of the Callus Analog approach . Fracture healing occurs primarily through the process of endochondral ossification , a process in which cartilage matrix is replaced by bone . This is the same process by which many bones are formed and grow . During endochondral ossification at the fracture site , chondrocytes express vascular endothelial growth factor , which induces vascular invasion of the cartilage matrix [49–51] . Along with the invading vasculature , osteoclasts that have entered the callus degrade the cartilage matrix . Previously , the chondrocytes were thought to undergo programmed cell death [52] , and concurrently , osteoblasts , which are delivered by the vasculature [19] , replace cartilage matrix with bone . In this two-stage theory , chondrogenesis—cartilage development—serves chiefly as a means for producing hypertrophic chondrocytes , and they , in turn , initiate bone formation ( carried out by other cells ) . However , a competing theory has emerged . Chondrocytes enter a transient stem cell-like state from which they transform into the osteoblasts and osteocytes that form the new bone [37 , 38] . The earlier theory envisioned chondrogenesis and osteogenesis as separate processes , whereas the more recent theory is mechanistically simpler: it envisions chondrogenesis and osteogenesis as characteristic sequential features of the same process . All of the wet-lab observations on which those two competing theories are based are below the resolution of Mechanism 4 . They are subsumed by the Fig 7 logic . So , at this stage , the model does not provide evidence for or against either theory . As we add new target attributes downstream , we envision replacing the Fig 7 logic with model mechanism details using the tuneable resolution process of Kirschner et al . [33] , which is a systems biology approach for discretized multi-level , multi-compartment computational models . The process involves fine- or coarse-graining of entities and activities . Such an approach allows for the adjustment of the level of resolution specific to a question , an experiment , or a level of interest . At that stage we should be able to challenge those competing theories . Because we are still early stage , the Mechanism 4-based simulated healing process comes with an ample supply of weaknesses and limitations . Both a weakness and limitation is that there are no ( fine-grain ) 1:1 counterparts to the cellular entities and molecular level events that are the focus of the majority of wet-lab experiments . As Callus Analog credibility improves and finer-grain features are included as validation targets , it will become feasible to increase model mechanism resolution further utilizing the tuneable resolution process in Kirschner et al . [33] . The day-10i illustration is a stage 1 requirement . It is an important source of information but also a source of uncertainty . A requisite for building an explanation for fracture healing is having staged representations of the same fracture ( e . g . , on days 4 , 7 , 10 , 14 , 20 , etc . ) that are , within reasonable tolerances , reliable , semi-quantitative , and scientific . There are currently no protocols to achieve that requisite . However , many histologists , pathologists , and biologists are trained in accurately illustrating representations of specimens , including tissue sections . A logical next step would be to acquire two ( or more ) independently generated day-10i illustrations of the same section and then document where and why they are similar and different . Thereafter , we envision protocols and methods for developing credible staged illustration representations becoming increasingly standardized , and , where feasible , automated . To increase standardization , we can draw from the robust best practices developed over decades for standardization of pathologic and histologic evaluations and reporting . We can also draw on the medical image registration methods [53] that enable rapid advances in computed tomography and magnetic resonance imaging . Feature discretization and simplification at stage 2 helps manage uncertainties , but the process itself is also a source of new types of uncertainty . There is a risk that increasing or decreasing grid mesh density can alter analog-to-mouse healing process mappings in scientifically meaningful ways . A good future time to assess that risk will be when Callus Analog insights have advanced sufficiently to begin exploring the first testable theory of fracture nonunion . The cellular components of all 80×80 μm areas within the day-7 tissue section Target Region are heterogeneous , but discretization requires that the corresponding analog grid space be occupied by one of nine TU types . To limit unintended bias , we can , as above , continue to draw from the mature best practices of pathologists to develop protocols to minimize added uncertainties . Longer term , we envision discretization protocols becoming automated . Near-term , several strategies can be explored to discover and ameliorate discretization weaknesses . Here are two examples . 1 ) Acquire two ( or more ) independently generated target region discretizations , and then independently develop a plausible explanatory analog for each that achieves the same final state Similarity criteria . 2 ) There will always be instances where it will be difficult for a domain expert or an automated process to make some TU assignment , such as choosing between a new marrow TU ( gray ) or an osteoblast-dominated TU ( burgundy ) , because those cell types are similar . In such cases , those grid spaces can be given a new designation , TU = g/b . At the start of each simulation experiment , all g/b spaces are randomly designated as either gray or burgundy . The result is a set of Monte Carlo Target Region starting states . One then cycles through the IR Protocol for each in parallel until the Target Region final state Similarity criterion is achieved . Both strategies require increased work , and automating IR Protocol tasks will help avoid reducing the overall workflow pace . Selecting an initial target region at stage 3 was essential to demonstrate feasibility . Moving forward , the approach must be expanded in stages to cover the entire callus , possibly as follows . First , develop and improve simulated healing processes for small portions of a callus and then explore how best to merge them incrementally to simulate more of the fracture healing phenomenon . It seems likely that multiple sub-callus processes will be needed to simulate healing within the entire callus . Evidence suggests that different callus subregions can be at somewhat different stages of repair and may progress at different paces . Separate simulations of independent subregions will help bring these issues into focus . A plausible next step would be to select a new day-7 target region ( possibly larger than 25×25 ) and determine if Mechanism 4 is able to achieve a corresponding day-10i final state with Similarity values ≥ 70% . Following that , we envision extending those two analog healing processes forward to day-14 and backward to day-4 . Each Mechanism 4 video is a sample from the circumscribed space of model healing processes , and each is biomimetic . Are all other Mechanisms in that space also biomimetic ? It is too early to answer , but it seems likely that the answer is no . Each video ( the record of one Monte Carlo trial ) provides a means to search for and address the emergence of non-biomimetic features . Observing more videos provides one with a better overall impression of the space of simulated healing . Domain experts observing videos can identify features that may be non-biomimetic , such as the small islands of teal and blue TUs within a gray/burgundy region in Fig 9A . Features that appear in one video may not appear in another . Assume that domain experts identify a likely non-biomimetic feature in several Mechanism 4 videos . Mechanism 4 would be falsified . We would then seek a marginally different—yet still parsimonious—model mechanism in which the logic used by each agent type has been revised to avoid exhibiting the non-biomimetic feature , while still meeting all similarity criteria . The revised Mechanism would circumscribe a smaller set of analog healing processes . In the preceding scenario , the videos provide domain experts with an entirely new means of thinking about callus healing . More broadly , simulated healing provides a new perspective on the actual healing process , and it is from that perspective that we encourage the use of simulations to enhance mechanism discovery . Doing so can help overcome translation barriers through the development of coarse-grained mechanism-based explanations . Later , as additional validation targets are met , incrementally better explanations will shrink the space of possible mechanism-based theories , and putative mechanisms will become finer-grained , which we anticipate will enable novel intervention strategies to be brought into focus . | Translation barriers have limited the generation of mechanism-based explanations of fracture healing processes . Those barriers help explain why , to date , biological therapeutics have had only a minor impact on fracture management . Alternative approaches are needed , and we present one that is intended to help develop incrementally better mechanism-based explanations of fracture healing phenomena . We created virtual Callus Analogs to simulate how the histologic appearance of a mouse fracture callus may transition from day-7 to day-10 . Callus Analogs use software-based model mechanisms , and simulation experiments enable challenging and improving those model mechanisms . During execution , model mechanism operation provides a coarse-grain explanation ( a theory ) of a four-day portion of the healing process . Simulated day-10 callus histologic images achieved 73–94% Similarity to a corresponding day-10 fracture callus image , thus demonstrating feasibility . Simulated healing provides an alternative perspective on the actual healing process and an alternative way of thinking about plausible fracture healing mechanisms . Our working hypothesis is that the approach can be extended to cover more of the healing process while making features of simulated and actual fracture healing increasingly analogous . The methods presented are intended to be extensible to other research areas that use histologic analysis to investigate and explain tissue level phenomena . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"biotechnology",
"traumatic",
"injury",
"medicine",
"and",
"health",
"sciences",
"engineering",
"and",
"technology",
"biomimetics",
"simulation",
"and",
"modeling",
"physiological",
"processes",
"histology",
"mathematics",
"statistics",
"(mathematics)",
"cartilage",
"bone",
"fracture",
"research",
"and",
"analysis",
"methods",
"bioengineering",
"tissue",
"repair",
"mathematical",
"and",
"statistical",
"techniques",
"connective",
"tissue",
"biological",
"tissue",
"monte",
"carlo",
"method",
"biophysics",
"critical",
"care",
"and",
"emergency",
"medicine",
"trauma",
"medicine",
"physics",
"anatomy",
"physiology",
"biology",
"and",
"life",
"sciences",
"physical",
"sciences",
"computational",
"biology",
"statistical",
"methods",
"biophysical",
"simulations"
] | 2018 | Simulation enabled search for explanatory mechanisms of the fracture healing process |
It has been hypothesized that schistosomiasis negatively influences immune reconstitution in people living with HIV starting antiretroviral therapy ( ART ) . In this study , we investigated the effect of schistosomiasis on the course of HIV infection in patients starting ART in a rural part of Tanzania . Retrospective study including patients prospectively enrolled in a HIV cohort in Ifakara , south-central Tanzania between January 1 , 2013 and April 1 , 2015 . Schistosomal circulating anodic antigen ( CAA ) was assessed in pre-ART cryopreserved plasma . Regression models were utilized to estimate the effect of CAA positivity on virological and immunological failure and a composite outcome of death/loss to follow-up ( LFU ) . At ART-initiation 19 . 1% ( 88/461 ) of patients were CAA-positive . A tendency of higher CD4 increases was seen in CAA-positive patients ( +182 cells/μl , interquartile range ( IQR ) , 87–285 cells/μl ) compared to CAA-negative patients ( +147 cells/μl , IQR , 55–234 cells/μl , p = 0 . 09 ) after 10 months of follow-up . After adjustment for baseline risk factors , CAA-positivity showed no association with virological or immunological failure . In CAA-positive patients , 22 . 7% ( 20/88 ) died or were LFU , compared to 29 . 5% ( 110/373 ) of CAA-negative patients ( hazard ratio ( HR ) : 0 . 76 , 95% confidence interval ( CI ) , 0 . 47–1 . 22 , p = 0 . 25 ) . After adjustment for age , sex , body mass index , educational attainment , WHO-stage , tuberculosis status , and year of ART initiation , CAA-positivity showed a trend of a decreased hazard of death/LFU ( HR: 0 . 58 , 95% CI: 0 . 32–1 . 05 , p = 0 . 07 ) , while CD4 count at baseline ( HR: 0 . 86 , 95% CI: 0 . 76–1 . 00 , p = 0 . 02 ) and MXD ( sum of eosinophils , basophils , and monocytes counts ) >1 , 100 cells/μl ( HR: 0 . 56 , 95% CI: 0 . 34–0 . 93 , p = 0 . 03 ) were identified as independently protective factors . Schistosomiasis is prevalent in this HIV cohort and may be beneficial for immunological reconstitution , while no effect on virological failure was apparent . A positive effect of schistosomiasis-induced immunomodulation on survival and retention in care needs confirmation in future studies .
The geographic distributions of HIV/AIDS and schistosomiasis largely overlap in sub-Saharan Africa , where Schistosoma prevalence reaches up to 30% in HIV cohorts [1–4] . In settings of coinfection , Schistosoma and HIV mutually interfere on several levels , which may impact the course of the associated diseases [5 , 6] . As both infections induce chronic modulations of the host’s immune system , interactions on this level are of particular interest [7] . Schistosomiasis and other helminth infections lead to an upregulation of T-helper cell type 2 ( Th-2 ) immune response and a downregulation of T-helper cell type 1 ( Th-1 ) immune response and of cytolytic activity of CD8 T-cells [8 , 9] . Properties of Th-1 immune response , which includes the secretion of interferon-γ and interleukin ( IL ) -2 by Th-1 lymphocytes promoting the activation of macrophages and dendritic cells and thereby enhances the ability to kill intracellular pathogens , are essential for the control of viral infections [10] . In line with these immunological findings , a study in an Ascaris lumbricoides-HIV coinfected population found higher levels of immune activation , HIV-RNA concentrations , and lower CD4 T-cell counts in individuals with Th-2 bias , as indicated by high A . lumbricoides fecal egg counts , eosinophilia , and IgE response , compared to patients with high A . lumbricoides fecal egg counts , low eosinophil count , and low IgE responses [11] . In a study carried out in Zimbabwe , a reduction in viral load with increase of CD4 T-cell count was seen in Schistosoma-HIV coinfected patients after antischistosomal treatment [12] , and similar effects have been shown after treatment of other helminth infections [13 , 14] . To our knowledge , only two studies have investigated the effect of Schistosoma coinfection on the course of HIV infection in people living with HIV ( PLWH ) under antiretroviral treatment ( ART ) [15 , 16] . Efraim et al . showed that in Schistosoma-HIV coinfected patients starting ART , the odds for immunologic treatment failure were four times higher and CD4 cell count increases were significantly lower compared to PLWH without concurrent schistosomiasis [16] . The effect of schistosomiasis on the virological response or on clinical outcomes was not assessed in that study . In settings of high Schistosoma prevalence and massive roll out of ART in HIV cohorts , Schistosoma-induced treatment failure would have major implications for ART programs . The working hypothesis of the current study was that Schistosoma coinfection has a negative impact on the course of HIV infection in patients starting ART . To test the hypothesis , we aimed to assess the effect of Schistosoma coinfection on patients’ response to ART in terms of ( i ) loss to follow-up ( LFU ) or mortality; ( ii ) immunological failure; and ( iii ) virological failure . The study was conducted in a well characterized HIV cohort in a rural part of south-central Tanzania and employed a highly sensitive diagnostic approach for schistosomiasis .
The Kilombero and Ulanga Antiretroviral Cohort ( KIULARCO ) received ethics approval from the institutional review board of the Ifakara Health Institute ( IHI ) and from the National Health Research Ethics Review Committee of the National Institute for Medical Research of Tanzania . All patients provided written informed consent at inclusion in the KIULARCO cohort and parents or guardians provided informed consent on behalf of all participants under the age of 18 years . We re-tested all circulating anodic antigen ( CAA ) -positive patients for Schistosoma infection . Patients with positive test results were treated with praziquantel ( 40 mg/kg twice at 4-week interval ) . This study was carried out in Ifakara , a primarily rural region of south-central Tanzania . Patients were recruited from KIULARCO , an ongoing , open , prospective observational HIV cohort of PLWH followed at the Chronic Disease Clinic in Ifakara ( CDCI ) . Further details of the KIULARCO cohort are given elsewhere [17 , 18] . Within the cohort , venous blood samples are drawn at routine clinic visits before , 3 months after ART initiation , and every 6 months thereafter . Plasma and cell pellets are cryopreserved at ˗80ºC and ˗20°C on site . The study was a retrospective analysis of cryopreserved plasma samples and of data collected within the existing HIV cohort . The study included all patients who were enrolled in KIULARCO , started ART between January 1 , 2013 and April 1 , 2015 , were older than 15 years , not pregnant , had a CD4 count performed at a maximum of 90 days before to 1 week after ART-start , and had at least one pre-ART plasma sample stored ( at a maximum of 30 days before ART-start ) . The Schistosoma infection status of each study participant at the moment of ART initiation was retrospectively assessed by testing 50 μl of pre-ART cryopreserved plasma for the presence of CAA with a lateral flow ( LF ) test with SCAA20 dry format [19 , 20] . Negative controls and CAA standard series indicated an assay threshold of 10 pg CAA per ml below which samples were designated as negative . If available , a second sample from 6–13 months after ART-initiation was tested for the presence of CAA . Of note , detection of active schistosomiasis has been simplified by recently developed assays for the detection of schistosomal CAA [19 , 20] . CAA originates from the gut of adult worms of Schistosoma mansoni and S . haematobium ( the two schistosome species endemic in Tanzania ) and is shed into the host circulation during active infection . After successful treatment with praziquantel , antigen levels decrease within days [21] . In addition , antigen levels are good indicators for worm burden . CAA is highly stable , and hence , can be detected in cryopreserved plasma samples for years after sample preparation [21–23] . Previously used enzyme-linked immunosorbent assays ( ELISAs ) as well as the currently used LF tests for detection of CAA in plasma are sensitive ( 80–95% ) and highly specific ( 98–100% ) for the diagnosis of active schistosomiasis , and the recently developed dry format of the latter facilitates usage in laboratories in endemic countries [20 , 24] . Because isolated eosinophil count in peripheral blood is not part of routine diagnostic procedures in the KIULARCO , we used MXD ( mixed cell count: sum of the absolute number of eosinophils , basophils , and monocytes counts ) as approximate value . Cut-off for elevated MXD value was set at 1 , 100 cells/μl [25] . The association between CAA-positivity and virological failure was assessed in a subgroup of study participants . Patients who were continuously CAA-positive ( CAA positive before ART-initiation , and CAA positive 6–13 months after ART-initiation ) were frequency-matched with two controls ( CAA-negative pre-ART and 6–13 months after ART-initiation ) for age , CD4 cell count , and tuberculosis status at baseline . In these patients , plasma HIV RNA levels were tested in a cryopreserved plasma sample drawn 6–13 months after ART-initiation . Plasma HIV RNA from 400 μl plasma was extracted using the NucleoSpin Virus kit ( Macherey-Nagel; Oensingen , Switzerland ) according to the manufacturer’s protocol . Viral RNA quantification was performed with the Brilliant III Ultra-Fast QRT-PCR Master Mix ( Agilent Technologies; La Jolla , CA , United States of America ) using the StepOneTM Real-Time PCR System ( Applied Biosystems; Foster City , CA , United States of America ) , with a detection limit of 60 viral RNA copies/ml of plasma . CD4 counts are routinely obtained after staining fresh whole blood samples with labeled antibodies: CD4 , CD3 , CD8 , and CD45 in TruCount tubes ( BD FACSCalibur; Franklin Lakes , NJ , United States of America ) . Data were extracted from a readily available electronic database of KIULARCO . The baseline was defined as the date of ART initiation . Continuous variables were summarized with medians and interquartile ranges ( IQR ) and categorical variables with frequencies and percentages . Association of CAA-positivity at ART-initiation with death or LFU was assessed with a multivariate Cox regression model . LFU was defined as no visit to the outpatient clinic for more than 6 months . The time of the event was defined as the day of the last follow-up visit documented in the database . Results were presented with hazard ratios ( HR ) and 95% confidence intervals ( CIs ) . The assumption of proportional hazards was confirmed by Schoenfeld’s global test ( p = 0 . 08 ) . The association of CAA-positivity at ART initiation with immunological failure was assessed with a multivariate logistic regression model . Immunological failure was defined as CD4 count falling below baseline or persistently <100 cells/μl at the time of the first measurement of CD4 cell count ≥6 months after ART-start . Results were presented with odds ratios ( ORs ) and 95% CIs . For both models the following variables were considered a priori as potential co-factors/confounders and were included in the multivariable model ( no variable selection was done ) : age ( divided by 10 ) , body mass index ( BMI ) , and CD4 count at baseline ( divided by 25 ) were used as continuous variables . Categorical variable included sex , educational attainment ( none and primary school vs . secondary school and college/university ) , and WHO clinical stages of HIV disease ( stage 1 and 2 vs . stage 3 and 4 ) . The variable “active tuberculosis” was defined as diagnosis of tuberculosis at baseline or during follow-up . The variable “MXD value” was dichotomized in normal vs . elevated ( >1 , 100 cells/μl , see above ) . The variable “year of starting ART” was added to the model used to assess the association of CAA-positivity with death/LFU . The variable “delay to CD4 testing” ( i . e . , time from ART initiation to measurement of CD4 cell count ) was added to the model used to assess the association of CAA-positivity with immunological failure . Linearity of the relationship between continuous explanatory variables and the log odds of the dependent variable was tested by adding polynomial terms of each continuous explanatory variable . Polynomial terms showing a significant association with the log odds of the dependent variable were included in the model . For the analysis of CAA-positivity as a risk factor for virological failure ( defined as HIV RNA concentrations above 1 , 000 copies/ml after 6–12 months of ART initiation ) , a logistic regression model was employed , controlling for the following potential co-factors/confounders ( defined as above ) : sex , BMI , educational attainment , WHO clinical stages of HIV disease , MXD value , and delay to HIV RNA testing . Results were presented with ORs and 95% CIs . Data were anonymized and analyzed using STATA version 12 . 1 ( StataCorp; College Station , TX , United States of America ) .
Among the 461 patients recruited , 88 ( 19 . 1% , 95% CI , 15 . 6–23 . 0% ) were CAA-positive . The median CAA titre at ART initiation was 163 pg/ml ( IQR , 27–760 pg/ml ) . After a median time of 36 weeks on ART ( IQR , 29–42 weeks ) , 36 remained CAA-positive , 17 became CAA-negative , and 35 had no follow-up plasma sample . In the group of 373 initially CAA-negative patients , 214 patients were continuously CAA-negative at ART-initiation and in the follow-up testing after a median time of 35 weeks ( IQR , 29–43 weeks ) , four became CAA-positive , and 155 had no follow-up plasma sample . Of 88 CAA-positive patients at ART initiation , 20 ( 22 . 7% ) died or were LFU , compared to 110 out of 373 ( 29 . 5% ) of CAA-negative patients ( HR: 0 . 76 , 95% CI , 0 . 47–1 . 22 , p = 0 . 25 ) . The median survival time in CAA-positive patients was 20 . 1 months ( IQR , 11 . 9–29 . 1 months ) , compared with 19 . 8 months ( IQR , 12 . 3–26 . 8 months; p = 0 . 57 ) for CAA-negative patients . Baseline risk factors for death or LFU are presented in Table 2 . After adjusting for age , sex , BMI , educational attainment , baseline CD4 count , WHO-stage , active tuberculosis , and year of ART initiation , CAA positivity showed a trend of a decreased hazard of death or LFU ( HR: 0 . 58 , 95% CI: 0 . 32–1 . 05 , p = 0 . 07 ) , while CD4 count and MXD >1 , 100 cells/μl ( HR: 0 . 56 , 95% CI: 0 . 34–0 . 93–0 , p = 0 . 03 ) were identified as independent protective factors ( Fig 1 ) . A follow-up CD4 count was available for 69 out of 88 ( 78 . 4% ) of CAA-positive and for 287 out of 373 ( 76 . 9% ) of CAA-negative patients . CAA-positive patients showed a median difference from baseline in CD4 of +182 cells/μl ( IQR , 87–285 cells/μl ) , compared to +147 cells/μl ( IQR , 55–234 cells/μl , p = 0 . 09 ) in CAA-negative patients . Median time from ART initiation to measurement of CD4 count was 37 . 9 weeks ( IQR , 32 . 1–43 . 7 weeks ) for CAA-positive patients , compared to 37 . 9 weeks ( IQR , 28 . 0–45 . 1 weeks ) in CAA-negative patients . MXD values had no effect on CD4 count changes ( MXD <1 , 100/μl: +151 cells/μl , IQR , 67–236 cells/μl; MXD >1 , 100/μl: +138 cells/μl , IQR , 51–221 cells/μl , p = 0 . 60 ) . Sixty-seven patients ( 18 . 8% ) met at least one WHO criterion for immunological failure . CAA positivity showed no effect on the risk of immunological failure ( OR: 0 . 78 , 95% CI: 0 . 39–1 . 59 , p = 0 . 50 ) . Baseline risk factors for immunological failure are presented in Table 3 . After adjusting for age , sex , BMI , educational attainment , WHO-stage , active tuberculosis , MXD , and median time from ART initiation to measurement of CD4 count , CAA positivity showed no association with immunological failure ( OR: 0 . 71 , 95% CI: 0 . 32–1 . 59 , p = 0 . 41 ) , whereas increasing CD4 count at ART initiation was an independent risk factor for immunological failure ( OR: 1 . 08 , 95% CI: 1 . 03–1 . 13 , p = 0 . 02 ) . One continuously CAA-positive patient and three continuously CAA-negative patients were excluded from the analysis due to non-availability of plasma samples for HIV PCR . After a median time of 43 . 4 weeks ( IQR , 38 . 0–49 . 9 weeks ) of ART , eight out of 35 ( 22 . 9% ) persistent CAA-positive patients had HIV RNA concentrations above 1 , 000 cp/ml , compared to 12 out of 67 ( 17 . 9% ) of continuously CAA-negative patients after frequency matching for tuberculosis-status , CD4 count , and age . After adjustment for baseline risk factors , positive CAA status showed no association with virological failure ( OR: 1 . 63 , 95% CI: 0 . 53–4 . 95 , p = 0 . 39 , Table 4 ) .
To our knowledge this is the first study analyzing the effect of Schistosoma coinfection on the immunovirological response and the long-term clinical outcome of PLWH starting ART . Highly prevalent active Schistosoma coinfection showed no association with virological and immunological treatment failure but a trend of decreased hazard of death or LFU in patients initiating ART in this mainly rural part of south-central Tanzania . Interestingly , an increase of the MXD value was identified as an independent protective factor against death and LFU in this cohort . The high prevalence of concomitant active schistosomiasis in this study , as determined by a highly sensitive assay ( i . e . , CAA ) and of schistosomal and helminth infection in general in other studies in PLWH corroborate the potential importance of helminth-HIV coinfection [11 , 16] . Our findings also highlight that schistosomiasis in adult patients is of special concern as large-scale treatment programs are often school-based and may exclude adults from access to appropriate therapy . An earlier study in KIULARCO yielded a prevalence of 43% for helminth and 11% of Schistosoma coinfection determined by stool and urine microscopy ( Cornelia Staehelin , unpublished data ) [26] . We do not believe that the higher prevalence of schistosomal infection in our study reflects a true increase of prevalence , but a considerably higher sensitivity of the technique of CAA detection in plasma compared to traditional stool microscopy-based diagnostics [24 , 26 , 27] . The low sensitivity of widely used microscopic diagnostics in schistosomal infection may be of special concern in settings were HIV co-exists . Schistosomal maturation and egg excretion are thought to depend of the host’s immune response and some studies suggest that HIV-induced CD4 cell depletion may be linked to a decreased luminal migration of schistosome eggs and an arrest of worm development [28–30] . From a clinical point of view , the implication of a possibly reduced sensitivity of microscopy-based diagnostic in HIV-infected individuals would be preclusion from anthelmintic treatment; in clinical trials on Schistosoma-HIV coinfection falsely negative tested Schistosoma coinfected patients would lead to misclassification and consequently to an underestimation of the effect under investigation . In such circumstances , the use of highly sensitive tests should be considered . In contrast to the findings of Efraim and colleagues , our study did not show an association of Schistosoma coinfection with an elevated risk for immunological failure or decelerated CD4 count gain in PLWH starting ART [16] . Considering evidence from previous research and pathophysiological mechanisms , a detrimental effect of helminth coinfection on CD4 counts recovery may be expected . Our findings , however , are in line with the findings of Muok and colleagues , who also showed a higher increase of CD4 count one month after initiating ART in patients who were infected with S . mansoni , compared to controls without Schistosoma infection [15] . Our results also suggest that elevated pre-ART CD4 cell counts are a risk factor for immunological failure . This counterintuitive association is a well-known phenomenon in European and sub-Saharan cohort studies and is generally explained by a closer clinical monitoring and better adherence of patients with lower CD4 cell count who are at an increased risk for opportunistic infections [31–33] . The rate of virological failure was substantial , which is a frequent problem , especially in rural settings of sub-Saharan Africa [34–36] . Essential factors for a virological successful ART are an adequate regimen and good adherence to therapy . High HIV RNA concentrations prior to initiation of ART are an additional risk factor for virological treatment failure [37] . Because helminth-induced immunomodulatory effects are associated with elevated HIV RNA levels in ART-naïve individuals , helminth infection could theoretically be associated with an elevated risk of virological failure [11] . Our study did not show an impact of schistosomal infection on virological treatment failure , probably because potential differences in HIV RNA concentrations prior to ART initiation are not important enough to affect efficacy of ART . The protective effect of schistosomiasis in PLWH starting ART in our study is counterintuitive . Possible explanations of the results include immunomodulatory properties of Schistosoma spp . , which may exert beneficial effects in the setting of ART-induced immunological reconstitution . Immune reconstitution inflammatory syndrome ( IRIS ) is a condition seen in PLWH with profound CD4 T cell depletion and is characterized by an overwhelming inflammatory response to pathogens due to a recovering T cell count after initiation of ART . IRIS is a common cause of early mortality of patients starting ART in sub-Saharan Africa [38] . Schistosoma-induced immunomodulation may have an attenuating effect on IRIS-related morbidity and mortality and by this means , improve the clinical outcome of PLWH starting ART . The statistically significant association of elevated MXD values with a reduced risk of death and LFU may support this assumption . MXD value is composed of absolute cell counts of monocytes , eosinophilic , and basophilic granulocytes , which are independent from HIV infection and disease status [39–41] . In the current setting in a primarily rural area of south-central Tanzania , elevated MXD values are most likely driven by helminth-induced eosinophilia and may be construed as a surrogate marker of helminth-induced immunomodulation in the HIV-infected host . Preceding studies in KIULARCO identified an important burden and variety of helminthic infections . Because immunomodulatory effects are a consistent feature of helminth infection , we believe that the weaker effect of schistosomal infection on survival/retention in care in our study may possibly be explained by competing immunomodulatory effects of other undetected helminths as our diagnostics in this study were focused on schistosomiasis . Indeed , about a quarter of PLWH without Schistosoma infection had elevated MXD values . Our study has several shortcomings that are offered for discussion . First , clinical data of KIULARCO are collected prospectively , but CAA was identified retrospectively in cryopreserved plasma samples with the immanent weaknesses of a retrospective study design . Second , due to a suboptimal ascertainment of mortality in our cohort , we employed a composite outcome of death/LFU , limiting our conclusions regarding the impact of schistosomiasis on mortality . However , LFU is a common issue in ART programs in resource-limited settings and mortality is inversely associated with the rate of LFU [42] . On the basis of an overall LFU of approximately 20% in KIULARCO , mortality of patients and LFU is expected to be about 50% [42] . Third , confounding factors cannot be excluded . In particular , schistosomiasis and elevated MXD values could be associated with other characteristics of the study population , which reduce mortality or LFU , remained unrecognized for which we could not control for ( e . g . , distance to the clinic or migratory mobility ) . Fourth , because pre-ART HIV-RNA concentrations were not assessed , it cannot be excluded that uneven distributions of HIV-RNA concentrations between both groups may have introduced bias in the presented results for the effect of CAA-positivity on the virological outcome . In conclusion , testing for CAA in plasma revealed a 19 . 1% prevalence of active schistosomiasis in PLWH starting ART in the KIULARCO in rural south-central Tanzania . This observation corroborates concerns of limited sensitivity of microscopy-based diagnostics , which might be additionally compromised in HIV-infected populations and may impede treatment for patients with a potentially life-threatening disease [43] . Our study could not detect detrimental properties of Schistosoma coinfection on immunological and virological response to ART but suggests that helminth-induced immunomodulatory mechanisms might enhance survival of PLWH starting ART . Our results need confirmation in future studies , which should consider ( i ) the dynamic aspect of HIV infection and therefore preferentially chose a longitudinal study design , including clinical outcomes; and ( ii ) that individuals are often coinfected by multiple parasitic worms that may affect the course of the HIV infection by comparable but interindividually variable immunomodulatory effects . To avoid misclassification , comprehensive , highly sensitive helminth diagnostics and markers of the host’s immunological response should be employed . Among these highly sensitive tests , the used CAA-test is a promising candidate and further use and developments may improve the acceptance as a future standard test . In the era of massive rollout of ART in sub-Saharan Africa and other tropical and sub-tropical countries , the issue of helminth-HIV coinfection may have major ramifications on the outcome of HIV treatment programs . Further research , especially in PLWH initiating ART , is urgently needed . | Infections with HIV and blood flukes ( Schistosoma ) both exert chronic modulatory effects on the host’s immune system . Coinfections , meaning the host is simultaneously infected with both pathogens , are common in sub-Saharan Africa . In this situation the induced immune modulation of one pathogen may affect the course of the disease induced by the other pathogen . One study showed that coinfection with Schistosoma in people living with HIV who begin antiretroviral therapy ( ART ) may have deleterious effects on the reconstitution of the HIV-induced immunosuppression . In the current study , we investigated the effect of Schistosoma coinfection on the recovery of the patient’s immune system , on the efficacy of ART to suppress HIV replication , and on a combined endpoint of lost to follow-up or death . We found that schistosomiasis may have beneficial effects on immune reconstitution , while no deleterious effect was detected on HIV-suppressive efficacy of ART . Surprisingly , our data suggest that schistosomiasis-induced immunomodulatory effects might be beneficial for survival and retention in care . Future studies are warranted to confirm these findings . In the era of increasing access to ART in sub-Saharan Africa , the issue of schistosomiasis-HIV coinfection may have major consequences on the outcome of HIV treatment programs . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"schistosoma",
"invertebrates",
"medicine",
"and",
"health",
"sciences",
"body",
"fluids",
"pathology",
"and",
"laboratory",
"medicine",
"helminths",
"antiviral",
"therapy",
"pathogens",
"immunology",
"tropical",
"diseases",
"microbiology",
"parasitic",
"diseases",
"animals",
"retroviruses",
"viruses",
"immunodeficiency",
"viruses",
"preventive",
"medicine",
"rna",
"viruses",
"antiretroviral",
"therapy",
"neglected",
"tropical",
"diseases",
"vaccination",
"and",
"immunization",
"public",
"and",
"occupational",
"health",
"infectious",
"diseases",
"medical",
"microbiology",
"hiv",
"microbial",
"pathogens",
"blood",
"plasma",
"immune",
"response",
"helminth",
"infections",
"schistosomiasis",
"eukaryota",
"blood",
"anatomy",
"viral",
"pathogens",
"physiology",
"co-infections",
"biology",
"and",
"life",
"sciences",
"lentivirus",
"organisms"
] | 2018 | Effect of schistosomiasis on the outcome of patients infected with HIV-1 starting antiretroviral therapy in rural Tanzania |
Fetal syncytiotrophoblasts form a unique fused multinuclear surface that is bathed in maternal blood , and constitutes the main interface between fetus and mother . Syncytiotrophoblasts are exposed to pathogens circulating in maternal blood , and appear to have unique resistance mechanisms against microbial invasion . These are due in part to the lack of intercellular junctions and their receptors , the Achilles heel of polarized mononuclear epithelia . However , the syncytium is immune to receptor-independent invasion as well , suggesting additional general defense mechanisms against infection . The difficulty of maintaining and manipulating primary human syncytiotrophoblasts in culture makes it challenging to investigate the cellular and molecular basis of host defenses in this unique tissue . Here we present a novel system to study placental pathogenesis using murine trophoblast stem cells ( mTSC ) that can be differentiated into syncytiotrophoblasts and recapitulate human placental syncytium . Consistent with previous results in primary human organ cultures , murine syncytiotrophoblasts were found to be resistant to infection with Listeria monocytogenes via direct invasion and cell-to-cell spread . Atomic force microscopy of murine syncytiotrophoblasts demonstrated that these cells have a greater elastic modulus than mononuclear trophoblasts . Disruption of the unusually dense actin structure – a diffuse meshwork of microfilaments - with Cytochalasin D led to a decrease in its elastic modulus by 25% . This correlated with a small but significant increase in invasion of L . monocytogenes into murine and human syncytium . These results suggest that the syncytial actin cytoskeleton may form a general barrier against pathogen entry in humans and mice . Moreover , murine TSCs are a genetically tractable model system for the investigation of specific pathways in syncytial host defenses .
Intrauterine infection is associated with pregnancy complications such as preterm labor [1] , which affects 10% of all live births [2] . All of the hematogenous placental microbes have at least partially intracellular life cycles [3] . Among these is L . monocytogenes , a facultative intracellular bacterium that causes foodborne disease in humans and other mammals . The relative risk of listeriosis is ∼115-fold higher in pregnant women compared to non-pregnant women of reproductive age [4] . The Centers for Disease Control reported 1 , 651 cases in the US during 2009–2011 , of these 227 ( 14% ) were pregnancy-associated [5] . L . monocytogenes triggers preterm labor and spreads to the fetus; the neonatal case-fatality rate is 22–45% [6]–[10] . Thus , pregnancy-associated listeriosis is a severe but rare disease . However , L . monocytogenes is ingested frequently by healthy adults [11] . Thus , it seems reasonable to hypothesize that the maternal-fetal interface forms an extremely effective barrier against infection . Perhaps the etiology of preterm labor is multifactorial even in cases of documented intrauterine infection . Indeed , recent evidence suggests that a combination of host genetic factors and bacterial products triggers preterm labor [12] . The placenta is a transient chimeric organ composed of maternal and fetal cells , and serves two major roles in the course of gestation: to nourish and to protect the fetus . The placenta must protect the fetus both from pathogens and from rejection by the maternal immune system [13] , [14] , which results in a unique immunological environment . The prevailing notion has been that fetal tolerance mechanisms create an immune-privileged site prone to infection [15]; however , recent evidence suggests that the placenta has effective innate defenses against microbial invasion and replication [3] , [16] , [17] . The placenta establishes its complex structure throughout the course of gestation: Fetal trophoblasts differentiate into several specialized cell types that perform critical placental functions [18] . Invasive trophoblasts penetrate the uterine lining ( decidua ) at the implantation site and remodel maternal arterioles to facilitate maternal blood flow into the intervillous space in humans or the labyrinth in mice . Inside these compartments , maternal blood bathes syncytiotrophoblasts ( or syncytium , SYN ) , which mediate nutrient and gas exchange between mother and fetus . Syncytiotrophoblasts form a multinucleated fused surface covered by a dense network of branched microvilli that spans an area of 12 m2 at the end of human gestation [19] . While the structural organization of the maternal-fetal interface differs between mouse and human ( villous versus labyrinthine placenta ) , the syncytium separates fetal and maternal blood in both species and serves analogous functions . We have previously shown that the main interface between mother and fetus , the human syncytium , is resistant to invasion by L . monocytogenes [20] . Further , we and others have found that the syncytium is resistant to viral and protozoan infection as well [17] , [21] , [22] . Therefore , we wanted to explore underlying mechanisms of syncytial resistance and identify factors that could make the syncytium more susceptible to infection , and possibly predispose to infection-triggered pregnancy complications . One underlying mechanism of its resistance to microbes is the paucity of receptors on its blood-bathed surface [23] and the lack of intercellular junctions; pathogens typically exploit receptors that are components of intercellular junctions to breach epithelial barriers [3] , [24] . However , we have also shown syncytial resistance to infection by receptor independent mechanisms including cell-to-cell spread of L . monocytogenes from infected macrophages , and direct invasion by T . gondii , suggesting that additional mechanisms contribute to syncytial defenses [20] , [21] . Syncytiotrophoblasts have an unusually dense cytoskeletal network most likely necessary to support such a laterally vast structure [25] . This led us to hypothesize that syncytiotrophoblasts form a physical barrier to invasion . The dense , branched microvilli on its surface may inhibit pathogen adhesion , and the unusually dense actin network could restrict the physical deformations necessary for pathogen entry . Primary human tissue can be difficult to obtain for many laboratories , is subject to donor-to-donor variability , and cannot be easily manipulated . Therefore , we turned to mouse trophoblast stem cells that can be differentiated in vitro into syncytiotrophoblasts and mimic the features of human syncytiotrophoblasts [26] . Isolation of mouse trophoblast stem cells ( mTSC ) is an established process that results in self-renewing cells that can be propagated in culture for many generations [27] . Recent advances in culturing mTSCs without underlying feeder cells [26] have made it suitable for study of placental infection with intracellular pathogens . Murine TSCs can be syncytialized in vitro by chemical inhibition of the ERK/MAP kinase pathway [26] , enabling experimental investigation of isolated syncytium in culture . Murine TSCs also open up the possibility to explore the contribution of specific molecular pathways to host defenses via isolation of mTSCs from genetically manipulated mice . We used syncytialized mTSCs from C57BL/6 mice to investigate the biophysical defense mechanisms of the syncytium . Like previous studies of the human syncytium , we found that the mouse syncytium is relatively resistant to infection by L . monocytogenes via two routes – direct invasion and cell-to-cell spread . Biophysical measurements of mouse syncytium with atomic force microscopy ( AFM ) demonstrated that it has a significantly greater resistance to deformation ( elasticity ) than mononuclear trophoblasts . Disruption of the actin cytoskeleton decreases syncytial elasticity and correlates with increased bacterial invasion . This finding was confirmed in primary human placental organ cultures , suggesting that the biophysical properties of the syncytium contribute to host defenses in both species .
Murine TSCs were infected with wild type L . monocytogenes , 10403S [28] at a multiplicity of infection ( MOI ) of 12 . Gentamicin was added to the culture medium at 1 hour post-inoculation ( p . i . ) to eliminate extracellular bacteria . On average 32% of mTSCs were infected with one bacterium at 2 hours p . i . , and we observed robust intracellular bacterial replication of L . monocytogenes in mTSCs ( Fig . 1 ) . The bacterial numbers increased by 64-fold , and the doubling time was 80 min between 2 and 8 hours p . i . Previous studies have shown that entry into primary human trophoblasts is dependent on the bacterial virulence factor Internalin A ( InlA ) [29] , [30] , and that the interaction of InlA with its host cell receptor E-cadherin is species specific [31] . Therefore , we also infected mTSCs with L . monocytogenes expressing murinized InlA ( InlAm ) , which was engineered for optimal interaction with mouse E-cadherin [32] , and with a mutant strain deficient in InlA ( del-InlA ) [30] ( Fig . 1A ) . Average InlAm-expressing bacteria per coverslip at 2 hours p . i . were 30-fold higher than for wild type ( p = 0 . 008 by Student's T-test ) . There was no difference in the degree of invasion for wild type versus del-InlA . Intracellular growth of each strain was similar in mTSCs with doubling times of 78–80 min between 2 and 8 hours p . i . These results are consistent with a role of InlA in direct invasion of mTSC , E-cadherin expression on mTSCs [26] , and species-specificity of InlA/E-cadherin interaction [31] . Next we characterized infection of murine syncytiotrophoblasts with L . monocytogenes . Differentiation of mTSCs into syncytiotrophoblasts was induced by the removal of FGF4 and heparin from the culture medium , and enhanced by addition of MEK inhibitor ( Fig . 2A ) [26] . Syncytiotrophoblasts were clearly recognized by their unique morphology , specifically the lack of long actin stress fibers and intercellular junctions that encompassed multiple nuclei ( Fig . 2B ) [25] , [26] . As previously described , short actin filaments formed a thick meshwork across syncytial patches . After 5 days under differentiating conditions , syncytiotrophoblasts covered 65% to 77% of the cell culture dish . The remaining area was occupied by mononuclear trophoblast cells that were not terminally differentiated; these contained stress fibers and prominent boundaries ( Fig . 2B ) . Differentiated TSCs were infected with InlAm-expressing L . monocytogenes . In addition , wild type L . monocytogenes was used to determine effects on invasion that are independent of E-cadherin expression on host cells . Because syncytialization of mTSCs in each well is incomplete , colony-forming units ( CFU ) would reflect the number of intracellular bacteria in both syncytium and neighboring mononuclear trophoblasts . Therefore , we determined the degree of infection at 2 hours p . i . by immunofluorescence microscopy . We outlined the area of syncytiotrophoblast ( Fig . 2C ) . The number of bacteria was represented by the green fluorescence intensity overlying the area of the syncytiotrophoblast in six random microscopic fields . Infection of mononuclear trophoblasts in the same well was determined by the same method , and the ratio of green fluorescence intensity overlying mononuclear trophoblast versus syncytiotrophoblasts was determined ( Fig . 2D ) . Invasion of syncytiotrophoblasts was lower than invasion of mononuclear trophoblasts by ∼25-fold ( Fig . 2D ) . These findings indicate that murine syncytiotrophoblasts are more resistant to bacterial invasion than mononuclear trophoblasts , and are consistent with previous observations in primary human placental organ cultures . Difference in invasion of mononuclear versus syncytiotrophoblasts was similar for wild type and InlAm-expressing L . monocytogenes , suggesting that resistance of the syncytium to pathogen entry is not just due to differential E-cadherin expression . We analyzed the surface resistance to deformation ( elasticity ) of syncytiotrophoblasts by microrheology with an atomic force microscope ( AFM ) [33] . For comparison , we measured the elastic modulus of mTSC . We chose undifferentiated mTSC instead of mononuclear trophoblasts , because the partially differentiated mononuclear trophoblasts are a heterogeneous population of cells at varying stages of differentiation towards syncytium . Given the large variability between samples in this system we decided to measure the extremes of the spectrum: undifferentiated mTSC versus terminally differentiated syncytium . Murine TSCs were plated onto chambered glass slides and differentiated for five days . Trophoblasts were probed with a polystyrene bead ( 5 um diameter ) mounted to a cantilever , whose deflection was detected optically ( Fig . 3A ) . Differences in elasticity of the syncytiotrophoblasts and mTSC were obtained by oscillating the cantilever and measuring the elastic response ( see Methods for details ) . The Young's modulus of elasticity of the cell cortex was then calculated from a Hertzian mechanics model . The median elasticity of syncytiotrophoblasts and mTSCs differed 4 . 8-fold ( 8 . 6 kPa versus 1 . 8 kPa ) ( Fig . 3B ) ; a difference that was statistically significant ( p = 4 . 7×10−5 by Student's T-test ) . The elasticity of mTSCs was within the expected range for mononuclear cells; for example , human embryonic stem cells have an elastic modulus of 1 . 2 kPA , while fibroblasts are typically stiffer with an elastic modulus of 3 . 5 kPA [34] . While a variety of cytoskeletal elements can affect cell elasticity , we hypothesized that the unique actin cytoskeletal organization of the syncytium contributes to its higher structural rigidity . Therefore , we wanted to investigate whether actin de-polymerization decreases syncytial resistance to deformation and increases susceptibility to bacterial invasion . We tested this hypothesis by measuring the elastic modulus of the syncytium before and after treatment with Cytochalasin D ( Cyto-D ) . Cyto-D depolymerizes the actin cytoskeleton [35] and has been shown to decrease the elastic modulus in other cell types [36] , [37] . For each independent experiment , three distinct spots of syncytium were measured by AFM prior to treatment; Cyto-D was then added to the culture coverslip for 40–60 minutes , and measurements were repeated in those exact spots . Treatment of murine syncytiotrophoblasts for 40–60 min with Cyto-D significantly decreased the median elastic modulus by 25% ( p = 0 . 001 by Student's T-test ) ( Fig . 3B ) . Others have shown that treatment of fibroblasts with Cyto-D can decrease cell stiffness by up to 50% [38] , [39] . Decreased elastic modulus in the syncytium was accompanied by consistent morphological changes ( Fig . 3C ) . The thick meshwork of smaller actin filaments was replaced by discrete puncta of actin aggregates . The modest effect of Cyto-D on the elastic modulus of the syncytium suggests that other cellular elements are important for maintaining syncytial stiffness . Colchicine , a microtubule polymerization inhibitor , did not have a significant effect ( data not shown ) on syncytial elasticity , implying yet other cytoskeletal or membrane features contribute to the tissue's rigidity . Next , we tested whether disruption of the cortical actin network in syncytiotrophoblasts resulted in higher rates of bacterial infection . We chose to investigate the effect of Cyto-D on syncytial infection via cell-to-cell spread because L . monocytogenes travels inside of leukocytes in maternal blood [40] , [41] . Fetal syncytiotrophoblasts are bathed in maternal blood . Hence , the syncytium comes in contact with infected maternal cells rather than extracellular bacteria . Syncytiotrophoblasts were treated with Cyto-D for 40–60 minutes , washed with PBS to remove Cyto-D , and co-incubated with infected murine macrophages in the presence of gentamicin . In this setup actin was disrupted only in the recipient trophoblast cells while bacterial protrusions formed normally in the donor macrophages . Because the physical force of a protrusion is enough for bacterial spread into an adjacent cell [42] , this system allowed us to investigate the defensive role of the syncytial actin network without impairment of protrusion formation in the donor cell . In most cell lines cell-to-cell spread begins as early as 4 hours p . i . [43] , [44] . Therefore we determined bacterial spread from macrophages into syncytiotrophoblasts at 5 hours p . i . by quantifying the number of bacterial foci in syncytiotrophoblasts ( Fig . 4A , B ) . A bacterial focus was included in the analysis when multiple bacteria were observed overlying syncytiotrophoblasts unbounded by the outline of a macrophage membrane; many such foci were surrounded by actin clouds . The number of bacterial foci was 2-fold higher in Cyto-D–treated syncytiotrophoblasts in comparison to untreated controls , ( p = 0 . 03 by Student's T-test ) ( Fig . 4B ) . In mononuclear trophoblasts Cyto-D removed stress fibers as well , but in contrast to syncytiotrophoblasts Cyto-D treatment did not significantly increase invasion into mononuclear trophoblast cells ( p = 0 . 33 by Student's T-test ) . In order to test whether these findings are relevant in humans we turned to primary human placental organ cultures [20] . It is not possible to measure the elastic strength of the human syncytium because we cannot maintain primary human syncytiotrophoblasts on glass coverslips—they require a gelatinous extracellular substratum . However , the actin cytoskeletal structure of the syncytium in primary human placental organ cultures is similar to mouse syncytiotrophoblasts ( compare Fig . 5A and 2B ) . Given the results in the mouse syncytium , we reasoned that decreased elastic modulus of human syncytium due to Cyto-D treatment could influence its susceptibility to bacterial invasion as well . Placental organ cultures were treated with Cyto-D for 40–60 minutes , washed with PBS to remove Cyto-D , and subsequently co-cultured with L . monocytogenes-infected human macrophages in the presence of Gentamicin to eliminate extracellular bacteria . Infection of the syncytium was quantified microscopically by bacterial co-localization with the beta-subunit of human chorionic gonadotropin ( b-hCG ) , a syncytial marker at 24 hours p . i . ( Fig . 5B ) [20] . Cyto-D treatment increased the number of bacteria co-localizing with b-hCG by 1 . 4-fold , a difference that was statistically significant ( Fig . 5C ) ( p = 0 . 04 by Student's T-test ) .
We adapted the mouse system of differentiated mTSCs to study placental defenses against infection . We show that multinucleated fused syncytiotrophoblasts have a greater elastic modulus than mononuclear trophoblasts , and that actin contributes to this phenotype . We present evidence that disruption of the actin cytoskeleton decreases the elastic modulus of syncytiotrophoblasts and increases bacterial spread into the syncytium in both mouse and human . Taken together , these findings suggest that the biophysical properties of the syncytium , a tissue unique to the placenta , may contribute to host defense mechanisms and protect the fetus . The cytoskeletal organization of the syncytium is characterized by an open lattice-like network of microtubules oriented in parallel to the syncytial surface , which support an apparently disordered mesh of actin microfilaments [25] , [45] , [46] . Our immunofluorescence microscopy showed such a diffuse actin and microtubule structure ( microtubule data not shown ) throughout the cytoplasm in both murine and primary human syncytiotrophoblasts [26] . Interestingly , erythrocytes have a similarly disorganized mesh of short actin fragments , which is necessary to withstand the shear forces they experience in the bloodstream [47] . The cytoskeletal organization of the syncytium may have evolved to withstand shear forces as well , since it is in direct contact with large volumes of maternal blood . The Young's modulus of elasticity was 4 . 8-fold higher in syncytiotrophoblasts in comparison to mononuclear mTSCs . Interestingly , small differences in the elastic modulus of different cell types or extracellular matrices have been shown to correlate with human disease states . Dulinska et al . , demonstrated that the Young's modulus in erythrocytes from patients with hemolytic anemia due to hereditary spherocytosis , thalassemia , or glucose-6-phophate dehydrogenase deficiency is 1 . 5 to 3 . 5-fold greater than the Young's modulus of normal erythrocytes [48] . A 5-fold increase in stiffness of leukemic cells correlates with clinical symptoms of leucostasis [49] , and a 2 to 4-fold increase in the stiffness of extracellular matrix leads to a small but significant increase in endothelial permeability and leukocyte transmigration , a process that occurs with aging and contributes to the pathogenesis of atherosclerosis [50] . We investigated whether the elastic properties of the syncytium correlate with susceptibility to infection . L . monocytogenes can infect non-phagocytic host cells by two mechanisms: direct internalin-mediated invasion and receptor-independent cell-to-cell spread . During cell-to-cell spread , L . monocytogenes propels itself in the cytosol of the donor cell into membrane protrusions ( listeriapods ) , which invaginate into and are taken up by neighboring cells . The force applied to the recipient cell by the listeriapod is estimated to be 0 . 03–0 . 3 nN [51] , and has been found to be sufficient for bacterial uptake in cultured cells [42] . The stiffness of the recipient cell may correlate with its susceptibility to bacterial transmission: L . monocytogenes secretes internalin C , which promotes cell-to-cell spread by binding to the cytosolic adaptor protein Tuba , which in turn leads to slackened cell-cell junctions and decreased cortical tension [52] . We used Cyto-D to disrupt the actin cytoskeleton of syncytiotrophoblasts , which decreased their elastic modulus by 25% . In our experimental setup , Cyto-D was added to the placental culture medium to disrupt actin structures and subsequently washed out before addition of untreated donor macrophages , allowing listeriapods in the donor cell to occur normally . Actin structures appeared to return to their original diffuse meshwork configuration over the course of the infection period ( data not shown ) , implying actin dynamics resumed with the removal of the drug . Exposure to Cyto-D led to a small increase in infection via cell-to-cell spread in both of our model systems: murine and human syncytiotrophoblasts . Our findings suggest two important conclusions: 1 ) the syncytial horizontal integrity is actively maintained by continuous actin assembly; and , 2 ) this structure is not permissive to the short-term perpendicular rearrangements engineered by listeriapods . We admit that the effect of Cyto-D treatment on elasticity and infection is small . Unfortunately , we were unable to decrease the elastic modulus even further by either increasing the dose or prolonging the exposure to Cyto-D , because of drug induced cytotoxicity ( data not shown ) . Thus we were unable to decrease the elastic modulus of syncytiotrophoblasts to a range comparable with mononuclear cells and we were also unable to maintain these conditions for the entire time of the experiment , which is dictated by the kinetics of cell-to-cell spread . These experimental limitations are one of the possible reasons why Cyto-D treatment led to only a small increase in infection . On the other hand , it is plausible that additional mechanisms influence the susceptibility of the syncytium to infection . Further , we cannot discount the possibility that Cyto-D has unknown side effects on placental cells that may affect their susceptibility to infection as well . Nevertheless , we speculate that syncytial elasticity influences its susceptibility to infection in vivo . Others have found that organization of the actin cytoskeleton can contribute to resistance against pathogen invasion . For instance , plant cells resist penetration by fungal pathogens via actin cytoskeletal reorganization [53] . On the other hand , it has been shown recently that some mammalian pathogens have evolved strategies to subvert the defenses of the actin cytoskeleton . The protozoan pathogen Toxoplasma gondii uses the virulence determinant Toxofilin to loosen the local host cell actin meshwork to facilitate invasion [54] . Pretreatment of host cells with jasplakinolide , which stabilizes actin filaments , renders them refractory to subsequent parasite entry [55] . However , Toxofilin is apparently inadequate to the task of syncytial actin rearrangement , as T . gondii are significantly inhibited from syncytial invasion [21] . The long-held view of the placenta as an immune compromised organ is slowly being eroded by the discovery of remarkable , unique innate defense mechanisms at the maternal-fetal interface . It appears that multiple forces have to coalesce to damage this barrier: placentas from preterm labor and congenital infections are often colonized by multiple pathogens [56] , [57] , and bacterial products synergize with viral infection to trigger preterm labor in the mouse model [58] , [59] . Damage to syncytiotrophoblasts by Plasmodium falciparum [60] , [61] or Trypanosoma cruzi [62] could lead to increased transmission rates of co-pathogens such as HIV [63] . Importantly , host genetic factors may also predispose to preterm labor triggered by bacterial products [12] . We suggest that even small changes in the integrity of the syncytial barrier may predispose the maternal-fetal interface to infections that lead to pregnancy complications , fetal damage and death . In summary , the placental syncytium is a unique structure that arose independently in many different mammals with hemochorial placentation [64] . Its critical role in the maintenance of healthy human pregnancy has been documented in terms of hormone production and proper nutrient and waste exchange . The contribution of its remarkable biophysical properties to resistance against pathogen invasion warrants further study .
This study was conducted according to the principles expressed in the Declaration of Helsinki . The study was approved by the Institutional Review Board at the University of California , San Francisco , where all experiments were performed ( H497-00836-28 ) . All patients provided written informed consent for the collection of samples and subsequent analysis . All chemicals were purchased from Sigma-Aldrich unless otherwise stated . Placentas from elective terminations of pregnancy ( gestational age 4 to 8 weeks ) were collected and prepared as previously described [20] . Briefly , fragments from the surface of the placenta were dissected into 1–3 mm tree-like villi , placed on Matrigel ( BD Biosciences , San Jose , CA ) -coated Transwell filters ( Millipore , Bedirica , MA , 30-mm diameter , 0 . 4 um pore size ) and cultured in Dulbecco's modified Eagle's medium-F12 medium ( DMEM-F12; 1∶1 , vol/vol ) supplemented with 20% fetal bovine serum ( FBS , Fisher Scientific ) , 1% L-glutamine and 1% penicillin/streptomycin ( Invitrogen , Carlsbad , CA ) . CellStart Humanized Substrate for Cell Culture ( Cell Therapy Systems ) was used for culture of mouse trophoblast stem cells ( mTSC ) . Cells were plated onto dishes pre-coated with CellStart diluted at 1∶20 with PBS , and maintained in RPMI-1640 with 20% FBS , 1% sodium pyruvate , 100 uM b-mercaptoethanol , 1% L-glutamine , 1% penicillin/streptomycin . FGF4 ( 25 ng/mL ) and Heparin ( 1 ug/mL ) were added fresh to media each time cells were thawed or split . For differentiation into syncytiotrophoblasts , mTSC were seeded onto coverslips pre-coated with 0 . 1% gelatin in 24-well dishes at 35 , 000 cells/well . Cells were maintained in RPMI-1640 with 20% FBS , 1% sodium pyruvate , 100 uM b-mercaptoethanol , 1% L-glutamine , 1% penicillin/streptomycin , and 10 uM U0126 ( MEK inhibitor , Pierce Biotechnology , Rockford , IL ) for 5 days . Fresh media containing MEK inhibitor was added to the culture every 2 days . The wild type strain of L . monocytogenes used in this study is 10403S [28] . L . monocytogenes with murinized InlA replacing WT InlA was a gift from Dr . Manuel Amieva [32] . For infections , bacteria were grown overnight to stationary phase in BHI ( Brain Heart Infusion broth ) at 30°C and washed once with PBS before dilution and infection . For mTSC infection cells were incubated in antibiotic-free medium for 1 hr before infection . 3×106 bacteria/mL were added for 60 minutes; cells were washed once with PBS and fresh media with gentamicin ( 50 ug/mL ) was added . At indicated times , cells were lysed , aliquots were plated on BHI agar plates , and CFU were enumerated . Five-day differentiated murine syncytiotrophoblasts ( mSYN ) were infected under the same conditions . At indicated times , cells were fixed , stained with phalloidin and polyclonal rabbit Listeria O antiserum and examined microscopically . Green fluorescence intensity in six random fields per area of syncytiotrophoblast versus mononuclear trophoblast was determined . For infection of mSYN via cell-to-cell spread from macrophages , J774 cell line ( ATCC TIB-67 ) was infected with 3×106 bacteria/mL ( MOI 3 ) for 60 minutes . Concurrently , mSYN was incubated with antibiotic-free media +/− Cyto-D ( 10 uM ) for 1 hr . Macrophages were washed 1× with PBS , gently scraped off the dishes and resuspended in mouse trophoblast media containing gentamicin ( 50 ug/mL ) . Infected macrophages were added to mSYN cultured on coverslips at 100 , 000 cells/well . Placental explants were infected via cell-to-cell spread from human macrophage-like U937 cells ( ATCC 1593 . 2 ) as previously described [20] . Briefly , U937 cells were grown in RPMI-1640 ( UCSF Cell Culture Facility ) containing 4500 mg/L glucose , 10% FBS and 1% penicillin/streptomycin ( Invitrogen ) . Forty-eight hrs prior to infection , cells were differentiated by addition of phorbol 12-myristate 13-acetate ( PMA; concentration 18 nM ) to the medium . On the day of infection , cells were incubated with antibiotic-free medium for 1 hr and subsequently infected with L . monocytogenes for 1 hr at an MOI of 3 . Concurrently , explants were incubated with antibiotic-free media +/− Cyto-D ( 10 uM ) for 1 hr , and subsequently washed 3× with PBS . U937 cells were washed once with PBS and lifted from culture plates by incubation in ice cold PBS without divalent cations and gentle scraping . U937 cells were resuspended in explant medium containing 50 ug/ml gentamicin , and 1×106 cells per transwell were added to the explants . Human placental explants were fixed in 3% paraformaldehyde , passed through a sucrose gradient and snap-frozen in OCT ( Ted Pella , Redding , CA ) . Histological slicing was performed on a Hacker-Slee cryostat . Glass slides with sections were incubated in acetone , soaked in blocking solution ( 1% bovine serum albumin ( BSA ) in PBS ) , then incubated with primary antibodies , rinsed in PBS , incubated with secondary antibodies , and affixed over Vectashield mounting medium with DAPI ( Vector Laboratories , Burlingame , CA ) . Mouse trophoblast cultures were fixed in 4% paraformaldehyde , blocked and permeabilized in 1% BSA and 0 . 1% Triton-X100 , then stained as described above in BSA/TritonX-100/PBS solution . Primary antibodies: polyclonal rabbit Listeria O antiserum ( 1∶1000 , BD Biosciences , San Jose , CA ) , monoclonal mouse anti-human b-hCG ( 1∶500 , Neomarkers , Fremont , CA , clone SPM105 . ) Secondary antibodies: Alexa Fluor 594 goat anti-mouse IgG ( 1∶500 , Invitrogen ) , Alexa Fluor 488 and 594 goat anti-rabbit IgG ( 1∶1000 & 1∶500 , Invitrogen ) . Alexa Fluor 594 – conjugated phalloidin ( 1∶100 , Invitrogen ) was used to stain for actin . Slides were viewed using an inverted TE2000-E microscope ( Nikon , Tokyo , Japan ) equipped with a 12-bit cooled CCD camera ( Q imaging , Surrey , Canada ) . Images were collected using Simple PCI software ( Hamamats , Sewickley , PA ) . Elasticity measurements were recorded at room temperature with a modified commercial AFM ( Bruker Bioscope Catalyst ) . Samples were mounted on a Zeiss Observer Z1 microscope . Optical images were acquired with an EM CCD camera ( Andor Ixon+ ) . Cantilevers were prepared by gluing ( Norland 61 ) polystyrene beads ( diameter 5 um ) to a tipless un-coated cantilever ( Veeco , customized MLCT , cantilever with a nominal spring constant of 0 . 01 N/m ) ; the cantilevers' spring constants were individually determined by the thermal vibration method before each measurement [65] . To monitor the elasticity of the syncytium membranes and stem cells , the polystyrene bead was moved in contact with the sample at a constant force of 0 . 5 nN in average . To measure the elasticity , a sinusoid oscillation with amplitude of 20 nm was applied to the cantilever vertically at 3 Hz . The amplitude and phase shift of the oscillatory cantilever deflection caused by viscoelasticity of the sample was detected with a lock-in amplifier ( Signal Recovery 7270 ) . The time constant of the lock-in amplifier was 1 second with the sensitivity at 500 mV . The elasticity measurements were recorded with a custom LabView program together with a National Instrument DAQcard ( PCI-6229 ) and an electronic signal filter ( Krohn-hite 3364 ) . For the analysis of the elastic modulus , calculations were based on the common Hertz model extended by Tu and Chen models [66] . The effect of Cyto-D on syncytial elasticity was observed by measuring the same spot of syncytium before and after 40–60 minutes of treatment . Images were prepared using ImageJ ( RSB , Bethesda , MD ) . JaCoP plugin for ImageJ was used for quantifying co-localization . Image files were blinded , and Manders' coefficients [67] were calculated , with individually set threshold for each channel in each image . The green channel ( LM ) threshold was set to exclude auto-fluorescence and background , but include all bacteria . Red fluorescence ( b-hCG ) threshold was set to include only the outer-most layer of cells on explant ( syncytium ) and exclude background . | Infection of the placenta can lead to pregnancy complications as well as fetal and maternal disease and death . We developed a novel system to study placental infections using murine fetal placental progenitor cells and the bacterial pathogen Listeria monocytogenes . In the mature placenta fetal progenitor cells fuse to form a large surface ( syncytium ) that is bathed in maternal blood and mediates nutrient and gas exchange between maternal and fetal circulation . We found that the syncytium resists physical deformation , and that its unusual cytoskeletal organization contributes to its elasticity . Weakening of its elastic properties correlated with increased susceptibility to infection . Our study presents a novel system to study placental infections , and provides new insights into the nature of the placental barrier . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [] | 2013 | Placental Syncytium Forms a Biophysical Barrier against Pathogen Invasion |
Rothmund-Thomson syndrome ( RTS ) is a rare autosomal recessive disorder characterized by skin rash ( poikiloderma ) , skeletal dysplasia , small stature , juvenile cataracts , sparse or absent hair , and predisposition to specific malignancies such as osteosarcoma and hematological neoplasms . RTS is caused by germ-line mutations in RECQL4 , a RecQ helicase family member . In vitro studies have identified functions for the ATP-dependent helicase of RECQL4 . However , its specific role in vivo remains unclear . To determine the physiological requirement and the biological functions of Recql4 helicase activity , we generated mice with an ATP-binding-deficient knock-in mutation ( Recql4K525A ) . Recql4K525A/K525A mice were strikingly normal in terms of embryonic development , body weight , hematopoiesis , B and T cell development , and physiological DNA damage repair . However , mice bearing two distinct truncating mutations Recql4G522Efs and Recql4R347* , that abolished not only the helicase but also the C-terminal domain , developed a profound bone marrow failure and decrease in survival similar to a Recql4 null allele . These results demonstrate that the ATP-dependent helicase activity of Recql4 is not essential for its physiological functions and that other domains might contribute to this phenotype . Future studies need to be performed to elucidate the complex interactions of RECQL4 domains and its contribution to the development of RTS .
Rothmund-Thomson syndrome ( RTS ) ( OMIM #268400 ) is a rare autosomal recessive disorder characterized by skin rash ( poikiloderma ) , skeletal dysplasia , small stature , sparse or absent hair , gastrointestinal complications , and high predisposition to specific malignancies such as osteosarcoma ( OS ) and hematological neoplasms [1 , 2] . RTS and the related RAPADILINO and Baller-Gerold syndromes are associated with damaging germ-line mutations in RECQL4 [3–5] . The majority of RECQL4 mutations are located in its helicase domain , yet the physiological role of this domain remains unclear [6] . Helicases are enzymes that unwind double-stranded or more complex DNA and RNA structures using energy from ATP hydrolysis . The unwinding of double-stranded DNA ( dsDNA ) is necessary to allow access to the DNA during replication , repair , recombination and transcription [7] . In humans , five members of the RecQ helicase family have been identified: RECQL1 , RECQL4 , RECQL5 , BLM and WRN . Mutations in three are associated with syndromes that present with premature aging and cancer-predisposition: WRN in Werner’s syndrome , BLM in Bloom’s syndrome , and RECQL4 in RTS [8 , 9] . RecQ-family helicases contain three highly conserved protein domains: an archetypal helicase domain , which contains seven conserved motifs that couple ATP hydrolysis to dsDNA strand separation [8 , 9]; the RecQ C-terminal ( RQC ) domain , which features a beta-hairpin motif , a winged-helix domain and a zinc-binding motif , for intervention in the binding of G quadruplex DNA and stabilization of DNA structures [10]; and the Helicase-and-ribonuclease-D-like C-terminal ( HRDC ) domain , which promotes stable DNA binding [11] . RECQL4 differs from the other family members as it has no HRDC domain and lacks a structurally conserved RQC domain . Instead , it contains the structurally unique domain called RecQ4-Zn2+-binding domain ( R4ZBD ) and , importantly , an N-terminal region of homology with the S . cerevisiae DNA replication initiation factor Sld2 [12 , 13] . Sld2 is an essential protein required for activation of replication origins in yeast [14] , and RECQL4 is the putative mammalian homologue . The human RECQL4 gene is located at the long arm of chromosome eight ( 8q24 . 3 ) and consists of 21 exons and 13 relatively short introns ( <100 bp in length ) , yielding a full-length transcript of 3 , 627bp [3] . Mutations in the RECQL4 gene have been found in the majority of RTS patients , and also in the related RAPADILINO and Baller-Gerold Syndromes [1 , 6] . Most mutations are either nonsense or frameshift mutations and are predicted to create truncated proteins [6] . Over half of these impact the reading frame at exons 8–14 causing abnormal translation and or truncation of the RECQL4 protein with the presumed loss of DNA helicase function [6] . Very few mutations are located in the N-terminal Sld2 region , and those that are reported are primarily silent or missense [6] . This finding has led to the hypothesis that the N-terminal domain is critical for organismal viability , and that inactivation of the helicase function is a critical effect of the mutation spectrum that is found in RTS . RECQL4 has been well characterized biochemically in terms of its single-stranded DNA ( ssDNA ) binding , ATP hydrolysis and DNA unwinding ability [15–17] . In vitro , RECQL4 ATPase and helicase activity is completely abolished by point mutations in the canonical Walker A and B motifs [18] . However , helicase-dead mutants can rescue cellular lethality in RECQL4-deficient DT40 chicken cells and murine hematopoietic cells in vitro [19 , 20] . In addition , mutations that affect the helicase domain not only affect its activity but can also lead to protein truncation or unstable proteins , which does not allow the specific assessment of the physiological requirement of the ATP-dependent helicase RECQL4 in isolation . To understand the biological functions of RECQL4 helicase activity in a whole animal context , we generated a mouse model with a knock-in point mutation that specifically abolishes its ATP-dependent helicase activity ( Recql4K525A ) and compared this to two different truncating mutations Recql4G522Efs and Recql4R347* . Here we show that mice with a specific deficiency in Recql4 helicase activity are strikingly normal , in contrast to pathogenic effects of truncating mutations that remove the entire helicase domain and downstream part of the protein .
We generated full-length wild-type mouse Recql4 protein , along with a K525A variant . The alanine substitution replaced a critical lysine , present in all Walker A motif-containing ATPases , that is necessary for ATP hydrolysis and corresponds to a previously analyzed human RECQL4 K508A mutation that lacks ATP-dependent helicase activity [15 , 21] ( Fig 1A ) . To biochemically characterize an ATPase deficient mutation in murine Recql4 , we established in vitro assays for Recql4 function . Both WT and K525A proteins had equivalent affinity for ssDNA binding , using an electrophoretic-mobility shift assay ( EMSA ) ( Fig 1B ) . In ATPase assays , the WT protein hydrolyzed ATP in a DNA-dependent manner whereas Recql4K525A showed no activity above background levels ( Fig 1C ) . Finally , we tested the ability of these recombinant proteins to unwind DNA , using a dsDNA substrate with a ssDNA loading site . If ATP-dependent helicase activity is present , the fluorophore-labeled strand is released from the quencher-labeled complementary strand providing a real-time fluorescent readout of unwinding activity ( Fig 1D ) . Robust ATP-dependent unwinding of the substrate was observed when using WT-Recql4 protein , whereas no change in fluorescence was observed using the Recql4K525A mutant consistent with an inability to unwind DNA ( Fig 1E ) . Together , our results demonstrate that the Recql4K525A protein is helicase and ATPase dead , despite equivalent protein stability and DNA binding properties . To understand the contribution of Recql4 helicase activity in the phenotypes of RTS , we generated an in vivo knock-in model of the K525A mutation . Sequencing of the Recql4 locus in targeted mice confirmed the change in nucleotides encoding lysine ( AAG ) to alanine ( GCA ) , and the resultant introduction of a unique MslI restriction enzyme site in the mutant allele ( Fig 2A ) . PCR amplification over the mutation site produces a 416 bp fragment in the wildtype that can cleaved to 361 bp ( +55 bp ) by MsiI when the mutation is present ( Fig 2B ) . Finally , to determine the expression and stability of the mutant protein in vivo we generated a rat monoclonal antibody against the first 200aa from the N-terminal of murine Recql4 ( clone 3B10 ) . The K525A mutant protein has the same predicted molecular mass ( ~133kDa ) as wild-type Recql4 and neither size nor abundance of the protein were affected by the mutation when assessed in thymocyte derived protein samples ( Fig 2C ) . Taken together , these results demonstrate that the loss of helicase activity does not affect the expression or stability of the Recql4 protein in vivo . We observed a slightly sub-Mendelian ratio of homozygous Recql4K525A/K525A animals at weaning from heterozygous breeding pairs , although this was not statistically significant by chi-squared test ( p = 0 . 6255 ) ( Fig 2D ) . Heterozygous and homozygous mice for the K525A mutation were viable and outwardly normal ( Fig 2E ) . Further , we observed no difference across genotypes or sexes in 10-week old animals assessed for body weight and body composition ( Fig 2F and 2G ) . Both male and female Recql4K525A/K525A animals were fertile and able to breed , and there was no difference in the survival when comparing the Recql4K525A/K525A and control genotypes using Kaplan Meier survival analysis ( Fig 2H ) . Collectively these results show that , unlike the embryonic lethality of Recql4 null alleles [20 , 22] , the Recql4K525A protein supports normal development and adult homeostasis . We previously reported that somatic deletion of Recql4 resulted in a fully penetrant bone marrow ( BM ) failure [20] . To determine the role of helicase activity in hematopoiesis , we assessed cohorts of adult wild-type , Recql4K525A/+ and Recql4K525A/K525A mice . Analysis of the peripheral blood ( PB ) revealed no changes in leukocyte or platelet numbers ( Fig 3A and 3B ) . The absolute red blood cell counts were subtly increased in both the K525A/+ and K525A/K525A mutants , however , the hemoglobin levels were not changed compared to the WT ( Fig 3C and 3D ) . Further analysis of the individual leukocyte subsets in the PB including granulocytes , macrophages , B lymphocytes , and CD4+ and CD8+ T lymphocytes revealed no significant differences in the proportions or absolute numbers of these populations ( Fig 3E ) . Within the BM , the total numbers of leukocytes were comparable across genotypes ( Fig 3F ) . Similarly , granulocyte and macrophage numbers were similar ( Fig 3G ) . Within the B-lymphoid populations ( B220+IgM- ) , the Pre-B , Pro-B , and Pre-Pro-B subpopulations were all unaltered in Recql4K525A/+ and Recql4K52A/K525A compared to WT littermates ( Fig 3H ) , as were the erythroid subpopulations ( Fig 3I ) . The frequencies and absolute numbers of the primitive hematopoietic stem cells ( HSCs ) and progenitors were also assessed . There were no major differences in the numbers of phenotypic long-term HSCs ( LT-HSC ) , short-term HSCs ( ST-HSC ) or multipotent progenitor ( MPP ( Lin- c-kit+ Sca-1+ CD34/Flt3 ) ) fractions , nor in their myeloid committed subpopulations ( Fig 3J–3L ) . There was an elevation of the phenotypic myelo-erythroid progenitor ( MEP , Lin-c-kit+Sca-1-CD34/FcγRII ) and colony forming unit-erythroid ( CFU-E , Lin-LKS-CD41-FcγRII-CD150-CD105+ ) in the Recql4K525A/+ animals , however the absolute change was very small and not statistically significant in the Recql4K525A/K525A mice . The basis for this mild elevation in RBC counts and committed erythroid progenitors is currently undefined . In summary , unlike for complete deletion of Recql4 , the abrogation of Recql4 helicase activity does not substantively perturb hematopoiesis in vivo . Our previous work demonstrated that the human helicase dead RECQL4 ( K508A ) was able to rescue in vitro B and T cell development from murine Recql4Δ/Δ hematopoietic cells [20] . To determine if this was also the case in vivo , T and B cell development was assessed from thymocytes and splenocytes respectively in 10-week-old mutant and WT mice . There was no difference in thymus cellularity or in the numbers of double positive CD4+/CD8+ nor the mature single positive CD4+ and CD8+ cells ( Fig 4A and 4B ) . Analysis of early intra-thymic progenitor cells ( double negative DN1-4 ) found no difference in the Recql4K525A/K525A compared to the WT littermates ( Fig 4C ) . In the spleen there was no significant difference in the cellularity or number of mature B cells ( Fig 4D and 4E ) . To determine the proportion of B cells in the follicular ( FO ) and marginal zone ( MZ ) of the white pulp of the spleen , we divided splenic cells with high expression of B220 and CD19 followed by analysis of CD21/CD23 expression ( FO B cells: CD21lowCD23high; MZ B cells: CD21highCD23low ) . No shift of B cells in the follicular or marginal zone compartments was apparent ( Fig 4F ) . Therefore , consistent with the prior retroviral rescue data in vitro , the ATP-dependent helicase function of Recql4 is not required for B or T cell development and homeostasis in adult mice . Given the role of RecQ helicases in repair of DNA damage [11] , we sought to determine if there was a function for the ATP-dependent helicase activity of Recql4 during physiologically induced DNA damage occurring following B cell stimulation . We isolated mature B cells from WT , Recql4+/K525A and Recql4K525A/K525A spleens and stimulated them in vitro using bacterial lipopolysaccharide ( LPS ) and rmIL-4 to induce proliferation and immunoglobulin gene ( Ig ) class switch recombination . The proliferation of the cells in response to LPS was the same irrespective of Recql4 helicase status as measured by cell-trace violet dilution kinetics ( Fig 4G and 4H ) . Additionally , the Recql4K525A/K525A cells underwent normal class switching to IgG1 ( Fig 4I ) . Therefore maturation , physiological activation induced proliferation , DNA damage and immunoglobulin rearrangement of B cells do not require the helicase activity of Recql4 . To further test the requirement for RecQ helicase activity we compared the response to DNA damaging agents in vitro . For these studies , we established GM-CSF dependent myeloid progenitor cell lines from R26-CreER Recql4fl/K525A and R26-CreER Recql4fl/+ ( control genotype ) by immortalization with HoxB8 [23] . To allow analysis of the requirement for RecQ helicase activity , the cells were treated with tamoxifen for 4 days to activate Cre-mediated deletion of the loxP flanked wild-type Recql4 ( Recql4fl ) allele . This resulted in the cells either becoming Recql4 heterozygous ( Recql4Δ/+ ) or expressing only the K525A mutant allele ( Recql4Δ/K525A ) . Isogenic cells from each genotype ( pre-treated with tamoxifen for 4 days or untreated ) were seeded in 96-well plates and the response to four different genotoxic agents was assessed: doxorubicin ( topoisomerase II inhibitor ) , hydroxyurea ( ribonucleotide reductase inhibitor ) , 4-nitroquinoline ( oxidative DNA damage ) and topotecan ( topoisomerase I inhibitor ) . Cell viability was measured after 48 hours using the CellTiter-Glo Luminescent assay . As shown in Fig 5A–5D , R26-CreER Recql4Δ/K525A had a comparable IC50 to Doxorubicin , Hydroxyurea , 4-Nitroquinoline , and Topotecan as the non-tamoxifen treated isogenic controls ( Figs 5A–5D and S3 ) . Additionally , R26-CreER Recql4Δ/+ showed a similar IC50 to Doxorubicin , Hydroxyurea , and Topotecan , except for 4-Nitroquinoline , which exhibited a mildly increased resistance in the tamoxifen-treated cells in comparison to the non-tamoxifen treated controls . These results indicate that the role of Recql4 in non-physiological , pharmacologically induced DNA repair does not depend on its helicase activity . The results so far establish that the ATP-dependent helicase activity of Recql4 is not essential in vivo . To directly compare the effects of having a protein truncating/hypomorphic allele compared to a helicase-dead full-length Recql4 protein , we established three additional point mutant alleles ( Fig 6A ) . The G522EfsX43 truncated mutation was generated as a co-incidental mutation during the Crispr/Cas9 mediate generation of the K525A allele and maps closely to the relatively common RTS associated mutations in human RECQL4 , S523TfsX35 and C525AfsX33 [6] . We also identified an N-ethyl-N-nitrosourea ( ENU ) mutagenesis induced truncated mutation R347* ( R355 in humans ) . Three RTS patients have been reported with p . Arg350GlyfsX21 mutations , mapping closely to this allele [6] ( S2 Fig ) . A second ENU induced mutation , M789K mutation ( V767 in human ) , was identified and used as a control for mutations within the Recql4 locus as this mutation was predicted to be benign . We crossed all the individual point mutant alleles ( M789K , K525A , G522Efs and R347* ) to the R26-CreER Recql4fl/fl line that we previously described and assessed the R26-CreER Recql4fl/PM in parallel with the R26-CreER Recql4fl/+ and R26-CreER Recql4fl/fl allele . [20] . At 8–10 weeks of age , mice were fed tamoxifen containing chow for 30 days to activate the Cre mediated deletion of the wild-type Recql4 floxed allele broadly throughout the body . Using this experimental design , the tamoxifen treated R26-CreER Recql4fl/fl animals ( completely Recql4 deficient ) developed fully penetrant BM failure [20] . The efficiency of Recql4 deletion was confirmed by PCR for genomic recombination ( S1 Fig ) . Analysis of PB ( Fig 6B ) showed an approximately 50% reduction in leukocytes and erythrocytes in mice expressing the truncating mutations G522Efs and R347* , very similar to mice rendered null for Recql4 protein expression ( Recql4Δ/Δ ) . The K525A and M789K mutation , as well as the heterozygous ( Recql4fl/+ ) control , did not show any significant change in leukocyte or erythroid indices . In addition , analysis of individual lineages within the PB showed a similar pattern across granulocytes , B lymphocytes , and CD4+ and CD8+ T lymphocytes in the truncating and null mutant ( Fig 6B–6D ) . When the BM was analyzed , leukocytes and erythroid precursors ( CD71+Ter119+ ) from R26-CreERki/+Recql4Δ/G522Efs and Recql4Δ/R347* mice showed a dramatic decrease , consistent with the BM failure phenotype we had previously described in Recql4Δ/Δ ( Fig 6E and 6F ) . The Recql4Δ/+ and K525A and M789K only expressing animals did not develop any phenotype after 30 days of treatment with tamoxifen and , whilst developing a pan-cytopenia in PB and BM , seven of eight G522Efs mice were still alive at the end of the treatment . In contrast , mice expressing the most severely truncating R347* allele developed a profound BM failure and three of five required euthanasia prior to 30 days of treatment , a phenotype similar to complete loss of Recql4 protein ( Fig 6G ) . Collectively , these analyses establish that the ATP-dependent helicase is not required for the physiological functions of Recql4 in vivo , however mutations resulting in truncated protein products are deleterious .
The N-terminal Sld2-domain of RECQL4 protein is unique among RecQ family members and has been shown to be critical for the initiation of DNA replication in chicken , Drosophila , Xenopus and human cells [12 , 17 , 19 , 24 , 25] , mostly likely through Sld2-domain dependent recruitment of the MCM10 and CTF4 factors to origins of replication [26] . The importance of the Sld2-domain to cell viability is reflected in the mutation spectrum detected in RTS patients–it is rarely mutated and always intact in at least one allele [6] . These findings implied that RTS and related disorders were caused by the loss of activity of the canonical ATP-dependent helicase domain , whose role in the replication initiation function is less clear . While initial in vitro studies using purified full-length human RECQL4 protein did not detect any unwinding potential on long DNA substrates [27] , it was later shown that RECQL4 could unwind shorter duplex regions if a single strand-loading region was provided [16 , 18] . This activity , together with the single-stranded DNA dependent ATP hydrolyzing activity , was lost in human Walker motif mutants RECQL4K508A and D605A . We now demonstrate that the murine Recql4K525A mutant protein is able to bind DNA but cannot hydrolyze ATP nor unwind DNA substrates , confirming that the mutant is helicase-dead and the homologue of human RECQL4K508A . Unexpectedly , however , we found that this protein behaved comparably to wild-type Recql4 in supporting viability , fertility and normal physiological development of mice . No phenotypes or symptoms consistent with RTS were observed . Perhaps more surprising , was the normal response of Recql4K525A/K525A cells in replication , and in their response to both physiologically or exogenously induced DNA damage . The most highly conserved domain of the RecQ helicases is the ATPase core . By analogy to multiple other DNA helicases ( reviewed in ref . [28] ) , it was assumed that the ATP-dependent helicase activity of RECQL4 is essential for normal cellular DNA metabolism . Several studies have reported findings consistent with this interpretation . Complementation experiments in Drosophila showed that the helicase-dead K898N mutant could not rescue the viability of RecQ4 null mutants [29] . Similarly , a helicase-inactive human D605A mutant could not restore replication of xRecql4 depleted cells in Xenopus egg extracts [12] . Murine studies to date have only assessed complete nulls or severe hypomorphic alleles . The deletion of the exons that precede the helicase domain ( exons 5–8 ) resulted in embryonic lethality by day 3–6 and a significantly truncated protein with no helicase or C-terminal domain [22] . When the entire helicase domain was deleted ( exons 9–13; in-frame deletion ) , mice were viable but showed high rates of perinatal lethality [30] and a truncated protein product of 480aa was predicted . With the aim to maintain an intact protein , a study replaced the last helicase coding exon ( 13 ) with a neomycin cassette . 95% of animals died within 2 weeks after birth and a substantial number of short transcripts covering exon 1 to 12 were reported , encoding potentially truncated products [31] . Since all these models effectively create proteins that lack the helicase and C-terminal domain , it is unclear if the observed phenotype could be attributed to the absence of the helicase domain only . Our study demonstrates that mice carrying the K525A helicase-dead mutation , with a stable full-length Recql4 protein , were viable and fertile with no apparent phenotype . The in vivo studies reported herein support a conclusion that the ATP-dependent helicase activity of Recql4 is not essential for replication or viability and that other domains account for these functions . A range of prior in vitro studies have pointed to the importance of the N-terminal Sld2-like region . Lethality of RECQL4-depleted chicken cells was rescued by expression of the N-terminal region only [19] . In addition , the N-terminal domain of RECQL4 was shown to physically interact with several proteins involved in DNA replication in Xenopus laevis and human cell extracts [32 , 33] . In our in vivo experiments however , we have shown that one copy of the N-terminal region alone ( R347* or G522Efs ) is insufficient for viability , indicating that a certain level of expression or localization of full-length Recql4 protein is required even if it has no ATP-ase or unwinding capacity . The high frequency of chromosome abnormalities found in cells from RTS patients and the increased cancer incidence rates , suggest that RECQL4 may have a role in maintaining genome stability through DNA repair [2] . Prior studies have attributed several DNA repair functions of RECQL4 to its ATP-dependent helicase region , but they also have noted participation of the N-terminal region . Lu et al . found that a helicase-dead K508M could not rescue the loss of DNA end resection and homologous recombination ( HR ) repair after RECQL4 siRNA knock down , suggesting that the ATP-dependent helicase function of RECQL4 is involved in HR [34] . However , they also showed that it is the N-terminus of RECQL4 that physically interacts with MRN and CtIP [35] . In a similar fashion , a role for RECQL4 in non-homologous end joining ( NHEJ ) was linked to its interaction with the Ku70-80 by the N-terminal domain [36] . RECQL4 deficiency has been associated with modulation of core proteins involved in base excision repair ( BER ) such as POLB , FEN1 , and APE1 . The latter has shown to specifically interact with the N-terminal region of RECQL4 [37] . Herein , we observed no differences in the sensitivity of helicase-dead mutant cells compared to WT cells in response to various kinds of DNA damage ( NHEJ , MMEJ , and HR for doxorubicin , BER for hydroxyurea , NER and BER for 4-nitroquinoline , and BER/SSBR for topotecan ) . Taken together there was no evidence for a role for the ATP-dependent function of Recql4 in the repair of pathological environmentally induced DNA damage . A recent study showed that RECQL4-depleted U2OS cells were also deficient in ATM dependent checkpoint activation in response to drug induced DNA DSBs . Complementation assays using helicase-inactive point mutants in Walker A ( K508G ) or Walker B motif ( D605A and E606A ) further indicated that this was the result of a lack of helicase activity [38] . The ATM pathway plays an equally important role in the physiological processes of DSB repair and recombination , such as V ( D ) J recombination in T cell development and class switch recombination in B cell activation [39–41] . In our in vivo analysis we did not detect any defect in T cell maturation at the CD4+/CD8+ double positive to CD4+ and CD8+ single positive transition or any earlier stage , nor did the mice develop any T cell lymphomas as a result of chromosomal anomalies . In addition , in vitro B cell activation and class switch recombination in helicase-dead splenic B lymphocytes was indistinguishable from that in WT cells , arguing that the helicase activity is not required for either physiological checkpoint activation or DNA damage repair . RTS patients , however , usually present with compound heterozygous mutations . It was reported that in 46% of RTS patients compound heterozygous mutations were present in the RECQL4 gene [6] . The majority of these mutations affect the helicase and C-terminal region and are predicted to create truncated proteins caused by an early stop codon , frameshift , or mis-splicing [6] . The phenotypes in RTS patients , although grossly similar , can vary widely in severity . The relatively common C525AfsX33 ( 12 alleles ) for example , has been found in all three syndromes ( RTS , RAPADILINO and Baller-Gerold ) and no single mutation has been assigned to a specific set of clinical features . In our study we demonstrate that mice carrying a sole truncating mutation ( G522Efs and R347* ) presented with a BM failure reminiscent of the Recql4 null . This was not seen in the helicase-dead K525A , the M789K or the WT heterozygous null mice . Furthermore , when we assessed BM , PB , thymus , and spleen from heterozygous and homozygous K525A helicase-dead mutants , we did not find any significant change . A similar observation was made for the human WRN helicase . A naturally occurring single nucleotide polymorphism ( R834C ) was shown to have less than 10% of the WT helicase activity , but normal exonuclease activity . None of the heterozygous or homozygous carriers of this mutation developed Werner Syndrome ( as defined by the clinical phenotype ) , clearly separating the WRN helicase function from other WRN functions [42] . Our findings demonstrate that helicase activity of Recql4 is also not required in vivo in mammals . Collectively , this study has demonstrated the ATP-dependent helicase activity of Recql4 is not physiologically essential for murine embryonic development or adult homeostasis , cellular replication and physiological DNA damage repair . However , mutations that create truncating proteins are not tolerated . Further studies will have to be performed to elucidate the complex interactions of Recql4 mutations , their role in RTS and the contribution of the individual Recql4 domains to its normal physiological function .
All animal experiments were approved by the Animal Ethics Committee , St . Vincent’s Hospital , Melbourne , Australia ( #007/14 and 011/15 ) . Animals were euthanized by CO2 asphyxiation or cervical dislocation . The full-length WT and K525A mutant codon optimized cDNA sequence of the mouse Recql4 containing 3xFLAG tag at the C-terminus were cloned into vector pFL-EGFP and transferred to the Multibac expression system to generate baculovirus [43] . Baculovirus infected High 5 cells were resuspended in TNG buffer ( 20mM TEA pH7 . 5 , 150mM NaCl , 10% glycerol , 1mM EDTA , 1x mammalian protease inhibitors ( Sigma-Aldrich ) and 1mM DTT . Mixtures were sonicated three times for 30 seconds on ice . Lysates were clarified by centrifugation at 50K x g 30 minutes . Anti-Flag M2–Affinity Gel ( Sigma-Aldrich ) resin was added and incubated for 60 minutes . Resin was extensively washed with TNG ( without PI or EDTA ) , then washed overnight in TNG containing Benzonase nuclease ( Sigma-Aldrich ) . Further washes were performed to remove the nuclease . Subsequently Recql4 was eluted with 100ug/ml Flag peptide in TNG . Helicase assays were performed according to Kaiser et al [18] . 80μl reactions containing 0 . 5μM protein and 50nM of a 15nt 3’-overhang ( OH ) DNA substrate in assay buffer ( 20mM HEPES pH 8 . 0 , 10mM NaCl , 5% Glycerol , 1mM MgCl2 , 0 . 5mM TCEP ) were assayed in an EnSpire 2300 microplate reader ( Perkin Elmer ) at 25°C . The DNA substrate ( T3-Cy3 annealed to B1-Dab ) contained a 3′-15nt polyT ssDNA loading site and a 15nt dsDNA part with a generic sequence . After recording baseline fluorescence for 60s ( Excitation 530nm / Emission 580nm ) , the helicase reaction was initiated by adding ATP to a final concentration of 1 . 25mM and the increasing Cy3-fluorescence as the quencher-labelled bottom-DNA strand is separated from the Cy3-labelled top-DNA strand , was recorded for 5 min . Measurements using H2O in place of ATP as well as reactions with buffer instead of protein served as blank and were subtracted from the ATP-data . PiColorLock phosphate detection reagent ( Expedeon ) was used to measure the presence of inorganic phosphate ( Pi ) release from ATP as a measure Recql4 ATPase activity . The proteins were assayed at 115nM in the presence of 1mM ATP and DNA in assay buffer ( 20mM HEPES pH 8 . 0 , 10mM NaCl , 5% Glycerol , 1mM MgCl2 , 0 . 5mM TCEP ) . Color change was measured at Abs650nm in an EnSpire 2300 microplate reader ( Perkin Elmer ) at 25°C . Electro-mobility shift assay ( EMSA ) was used to measure the relative binding of WT versus K525A mutant Recql4 to DNA binding . Protein was serially diluted from 300nM to 19nM and bound to 25nM single stranded DNA oligo XOm1 conjugated to IRDye680 . Bound Protein-DNA complex was separated on a 6% TBE/Acrylamide gel . The gel was directly imaged on a Li-Cor Odyssey CLx near-infrared fluorescence imaging system ( Millennium Science ) . Recql4K525A and Recql4G522Efs mice were generated using Cripsr/Cas9 methods by the Mouse Engineering at Garvan/ABR ( MEGA ) services ( Garvan Institute , Darlinghurst , Australia ) . Lysine 525 was mutated to Alanine ( AAG>GCA ) in single cell C57Bl/6 embryos via sgRNA-directed gene targeting and homologous recombination with a single stranded DNA oligonucleotide substrate . Viable pups were screened by DNA sequencing and one C57Bl/6 male carrying the K525A mutation on one allele and a 2bp insertion ( GA ) after the T521 codon ( G522Efs ) on the other allele was identified as a founder . The chemically ( ENU ) induced Recql4M789K and Recql4R347X mutations were obtained from the Australian Phenomics Facility ( APF , Canberra , Australia: IGL01381 and IGL01809 ) . Recql4fl/fl mice ( C57BL/6-Recql4tm2272Arte ) have been previously described [20 , 44] . Rosa26-CreERT2 mice on a C57Bl/6 background were purchased from The Jackson Laboratory ( B6 . 129-Gt ( ROSA ) 26Sortm1 ( cre/ERT2 ) Tyj/J , Stock Number: 008463 ) and have been previously described [20] . All lines were on a backcrossed C57Bl/6 background . ENU mutants were outcrossed at least 6 times and assessed across multiple generations to eliminate effects of any additional mutations . Tamoxifen containing food was prepared by Specialty Feeds ( Perth , Australia ) at 400mg/kg tamoxifen citrate ( Sigma Aldrich ) in a base of standard mouse chow . Genotyping of the K525A mutants was performed by PCR using the following primers: mRecql4 K525A MO36-F9: 5’-TAGACCTTATGAAACCTCAAAGCC-3’ and mRecql4 K525A MO36-R3: 5’- AGAACATTGGGCATTCGGC-3’ to yield a 591bp product , which was then digested with MslI ( NEB ) restriction enzyme to generate two fragments of 416 and 175bp for the WT or three fragments of 361 , 175 and 55bp for the K525A mutant . Primers for the M789K mutants are: mRecql4 M789K 1F: 5’- AATAGGTGGCAATGGGCAGG-3’ and: mRecql4 M789K 1R: 5’-GCACTCGGCGAAAGGATACA-3’ yielding a 420bp PCR product , uncut by MslI when M789K mutant , but cut in two ( 277 and 143bp ) when WT . The presence of the G522Efs and R347X mutations was determined by KASP ( competitive allele specific PCR ) technology ( LGC ) with custom designed ( G522Efs ) or facility provided ( R347X primer: 5’- GAAGGTGACCAAGTTCATGCTAAAGCGTTTGTTTTTCATGTTGAGTCG-3’ , 5’- GAAGGTCGGAGTCAACGGATTCAAAGCGTTTGTTTTTCATGTTGAGTCA-3’ , reverse primer 5’-GCTTCCCTAGACAGAGGGAACTATA-3’ ) sequences according to manufacturer instructions . Thymocyte lysates were prepared in RIPA buffer ( 50mM Tris , 150mM NaCl , 1% NP-40 , 0 . 5% sodium deoxycholate , 0 . 1% SDS , pH8 . 0 ) plus Complete protease inhibitor ( Roche , 04693116001 ) and PhosStop ( Roche , 4906837001 ) tablets . 25μg whole protein extracts were fractionated on pre-cast NuPAGE BOLT 8% Bis-Tris polyacrylamide gels ( Invitrogen ) and transferred onto Immobilon-P PVDF membranes ( Merck Millipore ) . Membranes were blocked with 5% milk in TBST and incubated overnight with rat monoclonal anti-mouse Recql4 antibody ( clone 3B10 , WEHI Antibody Services , Walter and Eliza Hall Institute Biotechnology Centre ) or mouse anti-β-Actin ( Sigma Aldrich , A1978 ) . Membranes were then probed with HRP-conjugated goat anti-rat ( Thermo Fisher Scientific , 31470 ) or anti-mouse ( Thermo Fisher Scientific , 31444 ) secondary antibodies and visualized using ECL Prime Substrate ( Amersham ) . Peripheral blood was analyzed on a hematological analyzer ( Sysmex KX-21N , Roche Diagnostics ) . For flow cytometric analysis , red blood cells were lysed using a red blood cell lysis buffer ( 150mM NH4Cl , 10mM KHCO3 , 0 . 1mM Na2EDTA , pH7 . 3 ) . Bones were flushed , spleens and thymus crushed , and single cell suspensions were prepared in PBS containing 2% FBS . Antibodies against murine Ter119 , CD71 , B220 , IgM , CD43 , CD19 , CD21 , CD23 , Mac-1 , Gr1 , F4/80 , CD4 , CD8 , TCRβ , CD25 , CD44 , Sca-1 , c-Kit , CD34 , FLT3 , FcγRII/III ( CD16/32 ) , CD41 , CD105 , CD150 , either biotinylated or conjugated with phycoerythrin , phycoerythrin-Cy7 , peridinin chlorophyll protein-Cy5 . 5 , allophycocyanin , allophycocyanin eFluor780 , eFluor660 or eFluor450 were all obtained from eBioscience , BioLegend or BD Pharmingen ( S1 Table ) [20 , 45 , 46] . Biotinylated antibodies were detected with streptavidin conjugated with Brilliant Violet-605 . 30 , 000–500 , 000 cells were acquired on a BD LSRIIFortessa and analyzed with FlowJo software Version 9 or 10 . 0 ( Treestar ) . B cells were purified from single cell spleen suspensions using a B Cell Isolation kit ( Miltenyi , 130-090-862 ) ; 106 cells per well were cultured in 6-well plates for 4 days in RPMI supplemented with 10% FCS , 100U/ml penicillin , 100ng/ml streptomycin , 2mM L-glutamine , 1 x MEM nonessential amino acids , 1mM sodium pyruvate , 50μM ß-mercaptoethanol , 15μg/ml LPS ( Invivogen , tlrl-3pelps ) and 10ng/ml recombinant murine IL-4 ( Peprotech , 214–14 ) , and stained with CellTrace Violet ( Thermo Fisher Scientific , C34557 ) and rat anti-mouse IgG1-APC ( BD Pharmingen , 550874 ) [47] . Stained cells were assessed using a LSRIIFortessa ( BD ) and data analysed using FACSDiva ( BD ) or FlowJo ( Tree Star ) software . Hoxb8 immortalized [23] R26-CreERT2 Recql4fl/+ ( control ) and R26-CreERT2 Recql4fl/K525A cells were maintained in IMDM , 10% FBS ( non-heat inactivated ) and 1% GM-CSF containing media ( BHK-HM5 cell conditioned media ) . The cells were treated for 4 days with 400nM 4-hydroxy tamoxifen ( Merck Millipore ) then genotyped to confirm complete recombination . Cells were then plated at 10 , 000 cells/well in 96 well plates ( Corning , CLS3610 ) and incubated for 48 hours with the indicated concentration of drugs in triplicates per dose ( dose range Doxorubicin: 0–0 . 5μM , Hydroxyurea: 0–0 . 5mM , 4-Nitroquinoline: 0–2μM and Topotecan: 0–0 . 5mM ) . Doxorubicin was obtained from St . Vincent’s Hospital Pharmacy . Hydroxyurea was purchased from Selleck . 4-Nitroquinoline and Topotecan were purchased from Sigma-Aldrich . Cell viability was measured using ATP-Lite ( Perkin Elmer ) as directed by the manufacturer and measured on an EnSpire plate reader ( Perkin Elmer ) . Data were plotted and the IC50 value calculated using Prism 7 software . The dose-response curve was plotted as mean±SEM . | DNA helicases unwind double-stranded nucleic acids using energy from ATP to access genetic information during cell replication . In humans , several families of helicases have been described and one of particular importance is the RecQ family , where mutations in three of five members cause human disease . RECQL4 is a member of this family and its mutation results in Rothmund-Thomson syndrome ( RTS ) . Prior studies have shown that defects in the helicase region of RECQL4 may contribute to the disease , but no studies have specifically assessed the biological effects of its absence in a whole animal model . In this study , we generated a mouse model with a specific point mutation resulting in a helicase-inactive Recql4 protein . We found that an absence of ATP-dependent helicase activity does not perturb the physiological functions of Recql4 with the homozygous mutants being normal . In contrast , when we assessed point mutations that generate protein truncations these were pathogenic . Our results suggest that the helicase function of Recql4 is not essential for its physiological functions and that other domains of this protein might account for its functions in diseases such as RTS . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"blood",
"cells",
"medicine",
"and",
"health",
"sciences",
"immune",
"cells",
"enzymes",
"immunology",
"enzymology",
"dna-binding",
"proteins",
"dna",
"damage",
"mutation",
"dna",
"white",
"blood",
"cells",
"animal",
"cells",
"proteins",
"t",
"cells",
"biochemistry",
"antibody-producing",
"cells",
"point",
"mutation",
"helicases",
"cell",
"biology",
"b",
"cells",
"nucleic",
"acids",
"genetics",
"protein",
"domains",
"biology",
"and",
"life",
"sciences",
"cellular",
"types"
] | 2019 | ATP-dependent helicase activity is dispensable for the physiological functions of Recql4 |
Kidney podocytes’ function depends on fingerlike projections ( foot processes ) that interdigitate with those from neighboring cells to form the glomerular filtration barrier . The integrity of the barrier depends on spatial control of dynamics of actin cytoskeleton in the foot processes . We determined how imbalances in regulation of actin cytoskeletal dynamics could result in pathological morphology . We obtained 3-D electron microscopy images of podocytes and used quantitative features to build dynamical models to investigate how regulation of actin dynamics within foot processes controls local morphology . We find that imbalances in regulation of actin bundling lead to chaotic spatial patterns that could impair the foot process morphology . Simulation results are consistent with experimental observations for cytoskeletal reconfiguration through dysregulated RhoA or Rac1 , and they predict compensatory mechanisms for biochemical stability . We conclude that podocyte morphology , optimized for filtration , is intrinsically fragile , whereby local transient biochemical imbalances may lead to permanent morphological changes associated with pathophysiology .
Podocytes , visceral epithelial cells of the kidney glomerulus , enable the selectivity of the glomerular filtration barrier through their specialized morphology . The cytoskeleton of each highly differentiated podocyte is composed of F-actin , microtubules , and intermediate filaments . All three of these cytoskeletal polymers form the cell body and primary processes , but only actin shapes the foot processes ( FPs ) , the delicate fingerlike projections that interdigitate to form the glomerular filtration barrier [1] . Actin is organized in a spatially specified fashion . Polymerization beneath the plasma membrane gives rise to a cortical actin network , and addition of crosslinkers result in the high density longitudinally aligned bundles , found in the center of the FPs [2] . The FPs establish contact between podocytes and the glomerular basement membrane , in addition to the cell-cell contact via specialized transmembrane junctions called the slit diaphragm , where plasma is filtered [3] . As a result of their highly dynamic filtration function , each FP must withstand tensile stresses ( due to glomerular expansion during systole ) and transverse shear stresses ( imposed by the fluid flow crossing the slit diaphragm ) , while maintaining contact with the neighboring cells as well as the basement membrane [4 , 5] . The loss of the characteristic FP morphology ( i . e . , foot process effacement ) and reduction in the number of podocytes are common hallmarks of chronic kidney disease ( CKD ) , whereby glomerular filtration becomes compromised and protein appears in urine ( i . e . , proteinuria ) [6 , 7] . The limited treatment options for CKD are in part due to the sensitive nature of these highly differentiated cells . Podocytes rapidly dedifferentiate after isolation of glomeruli , losing expression of key slit diaphragm proteins and the specialized morphology within 8 hours , and fully reverting to amorphous epithelial cell morphology within 48 hours [8] . Cultured primary or immortalized podocytes fail to fully differentiate [9] . These observations suggest that the mechanical stresses experienced by the podocyte in the mammalian glomerulus may regulate the integrity of the actin cytoskeleton within the FPs . Without the in vivo mechanical stimuli and the final shape signals , cultured podocytes present geometric characteristics ( e . g . , surface-to-volume ratio , eccentricity , characteristic length , etc . ) that are clearly different from the in vivo structure [10 , 11] . Consequently protein localization , gene expression levels , and active signaling pathways in cultured podocytes are not similar those of the in vivo cells [12] . Slit diaphragm has not yet been reconstituted in culture; hence , the unique localization of podocyte specific proteins that lead to the pro-differentiation signaling landscape and the stable actin cytoskeleton of the FPs is absent in cultured podocytes in vitro [13 , 14] . To understand how the maintenance of the FP morphology and slit diaphragm integrity are regulated , it is important to study the balance of signals within the context of the in vivo cell morphology . Rho GTPases play important roles in the regulation of the actin cytoskeleton [15] . Rac1 promotes the formation of a branched actin network , as found in lamellipodia whereas , RhoA promotes the formation of stress fibers and actin bundles . Expression and activity of both of these GTPases are tightly regulated in the healthy podocyte [16] . For example , while Rac1 knockdown may prevent protamine sulfate-driven FP effacement ( a standard animal model of acute podocyte injury ) , suggesting increased FP stability , Rac1 knockout animals that are subjected to chronic hypertension exhibit FP loss , proteinuria and glomerulosclerosis , showing an opposite effect [17] . Proteinuria and focal FP effacement are also observed when Rac1 is hyperactive [18] . It is not clear how an actin polymerization signaling hub , such as Rac1 , may act both as a stabilizing and destabilizing factor under these varying mechanical conditions . It is also puzzling that similar phenotypes could be observed for both Rac1 knockdown and overexpression . A similar outcome is observed with RhoA activity . In vivo induction of constitutively active or inactive RhoA damages the actin cytoskeleton causing loss of foot processes and proteinuria [19 , 20] . These findings indicate that the levels and activities of both RhoA and Rac1 need to be within a defined range to maintain morphological integrity of the foot processes . What is the quantitative relationship between local morphology and the GTPase regulators of biochemical and biophysical reactions underlying actin cytoskeleton dynamics that stabilize podocyte FPs ? To answer this question , we need to develop computationally tractable models based on realistic in situ morphologies . We developed dynamical models that combine upstream GTPase signaling with mechanical forces that control the podocyte actin cytoskeleton . Our models account for the actin stoichiometry , exchange between monomeric , filamentous , or bundled states , and the in vivo podocyte morphology . To account for the spatial specificity of the complex 3-D geometry of the FPs in vivo , we constructed a new quantitative model of a representative podocyte using 3-D serial blockface scanning electron microscope ( SBEM ) imaging in healthy rats . Our 3-D reconstructed model has sufficient resolution to capture the relationship between volume , surface area and characteristic length of the podocyte geometric features as well as the underlying biochemical and biophysical reactions . Our cytoskeleton model and the spatially specific simulations provide a mechanistic understanding of the in vivo observations regarding Rac1 and RhoA dynamics in podocytes , and their relationship to the cytoskeletal stability of FPs . The spatial simulations reveal the emergence of chaotic spatial heterogeneities within the actin cytoskeleton when Rac1/RhoA balance is altered providing a multiscale mechanism for foot process effacement due to propagation of chaotic behavior across FPs . Using this dynamical model , we also show how compensatory mechanisms could impact podocyte cytoskeletal integrity when they appear at different stages of regulatory dynamics .
We reconstructed the in situ 3-D geometry of interacting podocytes that would represent core spatial features of a podocyte that permits spatial modeling . SBEM image stacks were processed by manual segmentation and Gaussian filtering to reconstruct podocyte geometry with arborized processes that included nanoscale cell-cell junctions at the foot processes ( Fig 1 , Supplementary S1 Video ) . These reconstructions revealed a complex configuration of primary branches emanating from the cell body ( Supplementary Material 1 ) . Branching angles and the extent of secondary branching for these processes varied; however , the length and number of processes were similar among different podocytes ( S1 Fig ) ; also , despite the large level of deviations in the shape and projection pattern for primary processes , other key geometric characteristics , such as volume , principal dimensions , and surface area of individual podocytes exhibited remarkably low variability between individual podocytes ( Table 1 ) . While these studies were in progress , another report used SBEM to reconstruct qualitative features of podocytes [21] . Our reconstruction focused on identifying quantitative spatial parameters that could be used in dynamical models . The results from the analysis of five healthy rat podocytes are summarized in Table 1 , with the steps required to achieve this analysis illustrated in Fig 1A–1D . We assigned quantitative geometric parameters for podocyte morphology into three distinct compartments based on their reaction-diffusion dynamics: cell body ( CB ) , major processes ( MP ) that includes primary and secondary processes , and FPs , denoted with appropriate subscripts , respectively . We also reconstructed the nucleus ( Nu ) although it was not used in the dynamical models . The volume , V , and surface area , A , without subscripts , represent the respective parameters for the whole cell . The CB , MP and FP volumes were derived from the segmented images . Different filters that were based on reaction-diffusion dynamics , allowed reconstruction of a whole cell ( yellow surface in Fig 1E ) and a “foot process-free cell” ( blue surface in Fig 1E ) . The volumetric properties were computed using Seg3D ( Center for Integrative Biomedical Computing at University of Utah , Salt Lake City , Utah , USA ) and Virtual Cell ( Center for Cell Analysis and Modeling at University of Connecticut , Farmington , CT , USA ) ( Supplementary Material 1 ) . Boundaries of the different spatial components of the cell were estimated by applying a stratification method that is analogous to a well-known heat transfer problem [22] . The primary processes of a podocyte have a large surface-to-volume ratio in comparison to its cell body . Therefore , a podocyte with uniform volumetric synthesis and diffusion of a given protein within its entire volume and with transport of protein to the extracellular space ( proportional to local intracellular concentration ) will equilibrate to a much lower concentration at the primary processes relative to the cell body ( Fig 1F and 1G , analog heat transfer problem of uniform heat generation with convective boundary condition ) . Our analysis revealed that the CB volumes range from 30 to 50% of the total cell volume . FPs correspond to 20% of the total cell volume but constituted nearly 60% of the surface area; the remaining 50% to 30% of the volume corresponds to MP . The distances between the center of the CB and branch points , or ends of the MP , were measured using the filament tool in Imaris ( Bitplane , Zurich , Switzerland ) , as shown in Fig 1H and 1I . Since the branching parameters such as process length , counts , and distances were widely variable , we utilized the analytical descriptors of segregated volumetric units to construct a representative podocyte . From this analytical geometry , we obtained relevant reaction-diffusion equation parameters ( length , surface area , and volume ) at the whole cell level . The application of symmetry allowed us to reduce the mesh sizes needed for numerical discretization and the computational cost . FPs emanate from major processes in a symmetrical fashion . Therefore , in our analytical geometry , the major processes are generated as halves , using the sagittal plane as a reflexive boundary condition . The FPs emanate perpendicularly to such a boundary . Since there was no signature branching pattern for these cells , we also imposed axial symmetry . Consequently , the analytically constructed geometry produces simulations of a full podocyte cell , but at a quarter the size ( Fig 1J ) . The representative geometry was initially built by taking into account only the cell body and major processes . Once the surface-to-volume ( see Table 1 ) and distances of 18 ± 6 μm average distance from centroid to branch point , and 39 ± 2 μm for the three furthest endpoints , satisfied the analysis described above , the FPs were added . The quarter of the cell ( CB plus MP ) had a volume of 420 μm3 and surface area 695 μm2 . After 233 FPs were added , the final volume was 530 μm3 and surface area was 1683 μm2 ( Fig 1J ) . Table 1 demonstrates that the volumetric properties of the constructed geometry correspond to a good representation of the analyzed experimentally imaged cells . The surface area of the constructed FPs is on the upper range of the analyzed values . This is likely a conservative estimate; the resolution of the acquired images is of the length scale of the FPs , and a loss of surface detail is expected . The non-spatial computational model described below only uses volumetric variables , and is not affected by this assumption . To allow us to understand how actin cytoskeletal dynamics must be regulated to maintain this extraordinary cellular structure , we built a minimal kinetic dynamical model that describes the exchange of actin between monomeric ( G-actin ) , filamentous ( F-actin ) and bundled states . While the model vastly reduces the complexity of the actual biochemical machinery underlying actin dynamics [23–25] , the equations and parameters are approximately related to the key mechanisms controlling polymerization and bundling . For example , a generic “bundling coefficient” is used rather than different expressions corresponding to individual molecular contributors of actin-associated proteins or crosslinkers for bundling . However , different parameters in the model can be related to the activity of Rac1 and RhoA , allowing us to use these as surrogates for the corresponding signaling pathways . Based on experimental observations [26] , the actin cytoskeleton in the FP maintains its morphological stability . We used this cytoskeletal model to determine how the balance of GTPase activity affects the spatial stability of FP actin cytoskeleton and consequently its morphology . Initially , we developed an ordinary differential equation ( ODE ) model . We then mapped the ODE model to the reconstructed podocyte geometry and solved the reaction-diffusion equations for the corresponding partial differential equations ( PDEs ) numerically . Fig 2A shows the relationship between each state and the nomenclature used for parameters . We focused on the three discrete states of the actin cytoskeleton: namely monomeric G-actin , filamentous F-actin , and bundles ( i . e . stress fibers ) , represented by the variables Ga , Fa , and Bu , respectively . The model is described by Eqs 1–3 . In Eq 1 , the parameter γf lumps polymer elongation activity ( γf*Ga*Fa ) and nucleation activity ( 2*γf*Ga*Ga ) , as both will produce Fa; for nucleation , the coefficient is necessary for mass conservation since dimerization will form a filament containing 2 actins . The assumption that γf can represent both elongation and nucleation does not impact the model analysis , as presented in greater detail in the supplementary material ( S2 and S3 Figs ) . The filament state is further enhanced by positive feedback in the first term on the right hand side of Eq 1 , which describes nucleation mechanisms that depend on pre-existing filaments , such as Arp2/3-dependent branched nucleation . This term lumps the polymerization triggered by phosphorylation of nephrin and focal adhesion signaling through Nck , N-WASp or Rac1 [27 , 28] , and represented by a Hill function . Increasing the Hill coefficient does not change the qualitative behavior of the system ( S3 Fig ) . Therefore the parameter αf is non-zero only within the volume representing FPs , where nephrin signaling is localized . A non-linear functional form is necessary to represent both nephrin and small GTPase Rac1 driven polymerization [29–31] . The denominator identifies the high sensitivity region ( the positive feedback term is only significant if Fa is at least of the same order of magnitude as the parameter k ) , and ensures that the effective rate constant is bounded . In Eq 2 , activation of RhoA and crosslinking factors result in the formation of actin bundles . In the first term on the right hand of Eq 2 , new bundles are formed by merging filaments while existing bundles grow by addition of filaments , both governed by αb . Actin monomers can also be added to filaments within bundles ( γb*Ga*Bu ) . Once a filament becomes bundled , more crosslinks are added and it is unlikely to simultaneously break all connectors between a single filament and the bundle it belongs to . However , there is dissociation of monomers either from depolymerization or biomechanical stress-related rupture , both for filaments and bundles; we represent these , respectively , as -βf*Fa and -βb*Bu [32 , 33] in Eqs 1 and 2 , respectively . Eq 3 describes mass conservation within the cell . There are filaments and bundles in all parts of the cell; but we assume that the Ga pool is not strongly affected by dynamics of turnover of the F-actin in the CB and MP , and that crosslinked actin bundles are strongly localized to the FPs [26] . Therefore , we concern ourselves with the FP actin dynamics , where the small volume may lead to large changes in Fa or Bu . Thus , we consider the species Fa and Bu only within the FPs , while Ga freely diffuses to the different regions of the cell , as indicated by the different volume domains for the integrals in the mass conservation equation . Since the exact concentrations of actin at each state in the podocyte are unknown , all of the concentrations in the above equations are set as non-dimensional entities , which would allow quantitative comparison of state spaces . It is known that the ratio of total amount of monomeric actin to polymerized actin in the healthy podocyte is about 1:2 [34] . Here , we use the term “bundle concentration” to represent the concentration of actin molecules in the crosslinked bundle state . The intersections between the nullclines ( lines identifying values of variables that result in zero time derivative ) for the variables F-actin and bundles reveal the equilibrium points for the system; increasing any of the parameters presented in Fig 2B will move each nullcline as indicated by the arrows ( also in S4 Fig ) . Bundles are able to sustain mechanical stress better than a filamentous network , and a minimum bundle density is expected to be necessary for the morphological integrity of the individual FP [35] . If the polymerization or bundling rates are not sufficient to overcome the turnover rates , the equilibrium point moves to the “effacement region” of the diagram that is demarcated with gray shading , where loss of FP integrity and specialized morphology ( i . e . , effacement ) is expected . As illustrated by the two trajectories in Fig 2C , under conditions of a parameter regime with a single stable point , the FPs are able to sustain strong bundles irrespective of the initial conditions , or perturbation . This diagram represents ideal circumstances for a healthy stable podocyte . In contrast , reduced filament formation ( due to decreased Rac1 activity , or defective Nck signaling ) creates a second stable equilibrium point ( Fig 2D ) . If the system moves to this new state , with negligible bundles , the FP morphological integrity will be lost ( since this point is located in the effacement region of Fig 2B ) . Another parametric set that may result in destabilization of the system and loss of FP morphological integrity is via increased bundling ( αb ) or decrease in the turnover of the bundles ( βb ) . Depending whether αf is high or low , it may be subject to a cyclic behavior or collapse , respectively ( Fig 2E and 2F ) . The results in Fig 2 correspond to a non-spatial model of actin dynamics , with FPs comprising 20% of the total volume ( Eqs 1–3 ) . The behaviors described by the model are consistent with the experimental observations for response to different activity levels of RhoA in vivo . A basal level of RhoA in podocytes ( known to activate myosin and promote bundling ) has been shown to be necessary for healthy glomerular function [20] . Weak bundling ( αb ) would move the equilibrium point towards higher concentration of F-actin and lower concentration of bundles , shifting the equilibrium point towards the effacement region shown in Fig 2B . In the non-spatial model , the reasons for the damage caused by hyperactive RhoA are not obvious . Our model suggests that RhoA hyperactivity may lead to an imbalance between bundling and depolymerization , where the bundles “consume” all the actin , and the minimum density of filaments required for the positive feedback is no longer achieved . Eventually , the filament density becomes too low and the bundle turnover surpasses its formation . Once this happens , the cytoskeleton may either be subject to temporary ( Fig 2E ) or terminal collapse ( Fig 2F ) . As the bundles collapse , more monomeric actin becomes available . If the positive feedback for filament growth is sufficiently strong , the filament density recovers and the bundle density increases , resulting in cycles of weaker and stronger bundle density . The instances of weaker bundles may be sufficient to make the FP more susceptible to effacement under increased stress or even under physiological conditions . This scenario is further explored by spatial modeling in the next section . A weaker positive feedback ( αf ) is the mathematical representation of a system where Rac1 is inhibited . The low αf-system in Fig 2D has a stable equilibrium point with moderate bundles . The presence of another stable point , which represents the collapsed cytoskeleton ( minimal bundles or filaments ) , demonstrates that this region of parameter space is not as robust as the system with a single stable point as shown in Fig 2C . In addition , sustained biomechanical stress ( e . g . , high blood pressure ) corresponding to increased βb in the model , moves the bundle nullcline as indicated in Fig 2B , further decreasing the bundle concentration of the stronger stable point . Our model agrees with the physiological observation that , under such conditions , morphological damage may be observed [17] . When we use a partial differential equation ( PDE ) -based dynamical model with the representative 3-D podocyte geometry shown in Fig 1J , we observed spatial heterogeneities in the oscillatory behavior ( Fig 3 ) . This gives rise to permanent localized loss of bundles , leading to effacement of FPs ( Fig 3A–3D ) . As monomeric actin is consumed , it must diffuse from its largest pool in the cell body to all FPs . The varying distance from this major G-actin source to the arrays of FPs , which act as individual G-actin sinks , produce gradients in G-actin , triggering asynchronies in the availability of G-actin at individual FPs . Thus , bundles are collapsing in some FPs , while being strengthened in others . Consequently , as some FPs “release” their pool of G-actin , those monomers are sequestered by neighboring FPs , which are reciprocally reinforced . Fig 3B and 3C show the bundle concentration at different times after αb is increased . After elimination of some FPs , the remaining ones are able to build stronger bundles , temporarily enhancing their ability to overcome stress . However , the system is still unstable and morphological changes would progress slowly . Examination of Fig 3C , in particular , shows that although this is a purely deterministic system , the spatial pattern of bundle loss among the FPs appears to be random . Furthermore , the precise value of the perturbation in αb will produce different patterns . For example , different αb will result in a different steady cycle amplitude range and frequency in the ODE model , which will be translated into a different pattern for the loss of synchrony in the spatial model . Such unpredictable behavior following a change in initial conditions is the hallmark of a chaotic system . Fig 3D shows how the asynchrony can be followed by permanent damage in some FPs . In some of the FPs , bundle concentrations drop permanently to 0 , this would correspond to FP effacement . This example illustrates the potential impact of hyperactive RhoA on the FP morphological integrity . Each podocyte is subjected to spatially varying mechanical and biochemical signals: it adheres to the basal lamina supporting a segment of the coiled glomerular capillary vessel , and it interacts with several other podocytes . The cell-cell and cell-basement membrane interactions may lead to enhanced Rac1 activity , downstream of transmembrane slit diaphragm protein nephrin or focal adhesions [28 , 36] . Using our spatial model , we can demonstrate how this could lead to local damage . The ODE system with parameters as in Fig 2C has a single stable solution . However , this model is only valid if all parameters are uniform among all FPs . To explore and analyze the effect of differential actin bundling activities , we enhanced the ODE model to include two compartments ( S2-S6 Eqs ) . The monomer Ga is assumed to diffuse infinitely fast , resulting in the same concentration in the two compartments . The fraction “FP1” of the FPs in compartment 1 , retain the original value of αf , while there is a localized increase in the positive feedback strength for the complementary fraction of FPs , “FP2” , in compartment 2 . Fig 4A shows the steady state response for bundle concentrations in such an asymmetric cell , with constant positive feedback values αf and αf+Δαf for the fractions of FPs labeled FP1 ( blue mesh ) and FP2 ( red mesh ) , respectively . As expected , for small values of Δαf , each fraction of FPs have a new concentration of bundles , weaker ( FP1 ) or stronger ( FP2 ) than when Δαf is 0 ( green line ) . As Δαf and FP2 fraction increases , the bundles in the region FP1 collapse . Also note that as FP2 fraction increases , the total amount of actin becomes a limiting factor , and the local strength for the bundles cannot reach the high values observed in small FP2 fractions . This plot also illustrates that even though the machinery for F-actin polymerization is ubiquitous , the regions of the cell that are able to gather stronger polymerizing factors outcompete the remainder of the cell for the primary resource ( G-actin ) , resulting in uneven distribution of F-actin and bundles . This result is consistent with the observation that in podocytes , regulation of the actin cytoskeleton is tightly controlled in the FPs [26] , where a high concentration of proteins that mediate actin polymerization are specifically localized [37] . Fig 4C–4E show the impact of a localized transient enhancement ( shown in Fig 4B ) of the positive feedback on region FP2 only ( 50% of the FPs ) . If the stimulus is weak , both FP1 and FP2 regions recover the uniform bundle strength ( blue and red lines , respectively , in Fig 4C ) . For moderate stimulus , FP1 fraction loses bundles permanently and FP2 reaches a new steady state , with stronger bundles ( Fig 4D ) . For very strong Δαf , the initial response is as expected , but at longer times , the bundles in FP2 collapse while the bundles in FP1 are enhanced Fig 4E ) . The phase-plane diagrams in S5 Fig help explain the switch . At time zero , either FP1 or FP2 fraction has the same composition regarding bundles and F-actin . With strong Δαf , FP2 fraction consumes G-actin in order to develop more and more filaments , while the production rate of filaments for FP1 becomes weaker than its turnover . At this point , the two regions of FPs are asynchronous , and their volume fractions and proportions of F-actin and bundles ( i . e . , their position in the phase-plane ) will determine which fraction “wins over” the available G-actin ( note the trajectories in Fig 2C ) . In summary , the steady state response to a localized transient peak in the polymerization positive feedback can either lead to both regions of the FPs to recover the original uniform bundle concentration , or lead to one of the stimulated ( FP2 ) or unstimulated ( FP1 ) regions to collapse ( S6 Fig ) . The same range of responses was observed in our spatial simulations . Consistent with the results in Fig 4C–4G , the spatial simulations revealed that localized transient increase in the positive feedback αf may lead to localized changes in bundles , both in the regions within and adjacent to the stimulus ( Fig 5 ) . Time zero has all FPs in steady state and at same bundle concentration ( as in Fig 3A ) . After time 40 , only the region FP2 ( as marked in Fig 5 ) is subject to a transient increase in the parameter for positive feedback αf , while all other FPs are subject to constant αf ( all the other FPs in highlighted region are FP1 , time series for the feedback parameter αf is shown in S7 Fig ) . The timeseries plot in Fig 5A indicates the bundle concentrations in FPs identified by the corresponding colored arrowheads . Because the availability of monomeric actin is diffusion limited , a gradient of responses is observed . Within the region FP2 , there are FPs that recover the original bundle concentration once the stimulus is removed ( e . g . , the FP indicated by the orange arrow head in Fig 5A ) . There are also FPs that collapse within the same region ( red arrowhead ) . Interestingly , it is the FPs near the boundary that permanently lose their bundles . Similarly , among the FPs in close proximity to FP2 ( yet with constant αf ) , some also permanently collapse ( blue arrowhead ) . The 3-D movie of Fig 5 is in Supplementary S2 Video . In the movie , the bounding box identifies the region FP2 . A second example is shown in Supplementary S3 Video , with a significantly larger region for FP2 , comprising all FPs within the bounding box . Next , we studied the impact of decreasing bundle turnover rate βb and the potential ways to regulate podocyte FP integrity . Bundle turnover rate , βb , accounts for depolymerization and damage of bundles due to mechanical stress in the glomerulus . It is expected that under high or low blood pressure , βb would be enhanced or attenuated , respectively . As described above , cytoskeletal stability is dictated by the relationship between the F-actin and bundles nullclines , and as illustrated in Fig 2B , different parameters move the stable equilibrium points in different directions . Initially , the healthy podocyte is in steady state in its single equilibrium point ( Fig 6A ) . By decreasing the bundle turnover rate , the system is destabilized and progressive damage and effacement of FPs are observed ( Fig 6B and 6C ) . Here in this complex geometry , the interplay of Eq 1 and Eq 2 , through diffusion of G-actin , produce marked spatial heterogeneity in the pattern of bundles . The trivial compensatory mechanism would be to restore the original value of βb after a certain amount of time Δt1 ( Fig 6D ) . Most of the FPs that survived up to the time of the correction ( Fig 6B ) would have reached stability after elapsed time Δt2 ( Fig 6E ) . A similar result is observed if the compensatory mechanism is applied at a later time ( S8 Fig ) . The permanent effacement of several FPs provides a larger pool of actin to be incorporated by the surviving ones , resulting in stronger bundles . This would correspond to the loss of some FPs and the strengthening of others . The timecourse for the bundles in the FPs highlighted by solid arrows is plotted with corresponding solid lines in Fig 6H . The gray arrowhead identifies the time point used for the 3-D snapshots . These spatially asynchronous and irregular timecourses further substantiate the chaotic tendencies of this system . A second alternative for a compensatory mechanism is the decrease of αb , representing decreased bundling coefficient or concentration of crosslinks ( Fig 6F ) . We show that this mechanism can potentially achieve equivalent success as the trivial case of restoring βb . Once again , the sooner the intervention , the smaller the population of damaged FPs ( Fig 6F and 6G , and timecourse of FPs identified by dashed arrows represented in dashed lines in Fig 6H and S8 Fig ) . We also explored the impact of increasing the positive feedback , αf , corresponding to hyperactivating Rac1 or other polymerization signals coming from the slit diaphragm ( Fig 6I and 6J ) . Now , high level of polymerized actin is maintained in each of the surviving FPs; however , there is a new mode of oscillation for bundles ( and F-actin ) along each FP . The simulations suggest that if applied early enough , this compensatory mechanism may help keep the FP morphology intact ( Fig 6J ) . Similarly to the previous cases , the surviving FPs are able to build stronger bundles , and in spite of the oscillatory behavior , equivalent number of FPs seem to be preserved in comparison to the previous examples ( Fig 6J , 6E and 6G ) . The overall conclusion is that there are several potential compensatory mechanisms that may be activated to restore FP stability , and the sooner the regulatory response is activated , the larger the number of surviving FPs .
We acquired 3-D electron microscope imaging data of rat kidney glomeruli , individually segmented podocytes , and quantitatively analyzed their morphological features to determine the average volume and surface area of the cell body , major processes , and foot processes . This resulted in construction of a representative 3-D podocyte model that is amenable to PDE-based reaction-diffusion modeling with physiologically relevant mechanisms . To the best of our knowledge , this is the first time such a detailed quantitative reconstruction was ever achieved . The variability in the geometrical parameters measured between the five cells of Table 1 was remarkably small . Through our morphometric analyses , we show that the FPs corresponded to 20% of the volume of the podocyte , while doubling its surface area . Consequently there is a large pool of diffusive monomeric G-actin that is available for the FPs . However , the G-actin pool is not instantly or uniformly available: the monomers must diffuse across the different regions of the cell . Our representative podocyte geometry carefully preserved the length-scales , volumetric and diffusive properties , quantity and distribution of FPs , primary processes , and cell body , thus enabling physiologically relevant simulations . The range of spatiotemporal dynamics within the real geometries of podocytes is likely to be broader than this idealized geometry , but we feel that this idealized geometry will be representative; it also has the advantage of being more computationally tractable , both in simulation time and for analysis of simulation results . To explore the consequences of the unique podocyte morphology on the maintenance of cytoskeletal integrity of the FPs , we developed a minimal model of actin dynamics and compared non-spatial ( ODE-based ) simulation results with full reaction-diffusion ( PDE-based ) models in the constructed geometry . The minimal model represents key regulatory mechanisms controlling actin polymerization and filament bundling . It contains positive feedback in actin polymerizations , representing Arp2/3-dependent branching nucleation on preformed mother filaments . This positive feedback results in regions of parameter space that produce oscillatory F-actin and bundle kinetics in the ODE simulations ( Fig 2 ) . Another mechanism that would be similarly represented by this positive feedback is actin polymerization downstream of nephrin phosphorylation and Nck localization to the FP slit diaphragm . However , it should be emphasized that we did not model the detailed biochemistry regulating actin polymerization and bundling , some of which is still unclear . Rather , our model uses a reasonable level of biochemical specification that allows us to focus on the influence of the complex geometry of the kidney podocyte . Spatial simulations explicitly consider the process of diffusion of G-actin in and out of FPs from the large reservoir of the cell body and primary processes . The interplay of this diffusion and the positive feedback inherent in the actin dynamics can produce sharp regional differences in the level of actin bundling within closely neighboring FPs . Without bundled actin , the FP would be resorbed into the parent process leading to effacement . Therefore , we assumed that the loss of bundles in FPs leads to loss of structural integrity . Even when a change in bundling activity is distributed uniformly throughout the cell , an apparently chaotic , spatiotemporal pattern of bundle formation and collapse is observed in the FPs ( Fig 3 ) . One of the major unanswered questions in podocyte biology is the source of the inherent heterogeneity in podocytopathies . Diseases , such as focal segmental glomerular sclerosis , affect only a subpopulation of podocytes with a large spatial variability in disease etiology . Based on our spatial dynamical model simulations , we suggest that even global changes in stress ( e . g . , due to hypertension ) can lead to selective effacement of FPs with large spatial variability , which may explain the pathophysiological disease progression of numerous podocytopathies . The balance of Rac1 and RhoA activity ( that is represented respectively by actin polymerization and bundling in the model ) needs to be tightly regulated to maintain FP stability . This is all the more critical in FPs , where these molecules may be present at low copy number and result in potential stochastic instabilities . Stress ( whether focal or global , transient or continuous ) can lead to spatially heterogeneous chaotic behavior , and ultimately , to irregular heterogeneous patterns of bundle collapse in FPs . Fig 4 shows how transient activation of Rac1 ( i . e . , actin polymerization ) in one collective region of FPs can shift the steady state balance between F-actin and bundles in all regions . A much more localized transient perturbation , as shown in the spatial simulation of Fig 5 , can produce permanent changes in the bundle distribution within the system , both within the perturbed region and its immediate vicinity . Unfortunately , experimentally accessible live podocyte models that retain this cell’s unique morphology are not yet available . When this experimental problem is solved , direct manipulation of the regulatory mechanisms ( e . g . using local photoactivation of RhoA or Rac1 [38] ) will help to test whether the chaotic responses predicted by our model can be validated . Direct damage to FPs , due to high blood pressure for example , can be captured in the bundling turnover parameter βb . Changing βb globally , but transiently , can produce permanent changes in bundle distribution ( Fig 6A–6C ) ; the severity of FP effacement ( represented by loss of bundles ) is directly related to the duration of the perturbation . It is possible that regulatory mechanisms could be modulated to alter Rac1 and RhoA activities in response to stress to ameliorate the imbalance . The behavior of the system under three potential compensatory interventions in response to lowered βb are explored in Fig 6D–6J . These , too , could be tested experimentally once an appropriate in vivo or culture preparation is established . The model for the actin cytoskeleton of the podocyte FP provides a framework for understanding recent findings on the in vivo activity levels of the small GTPases RhoA and Rac1 . Our model captures the need for a balance between polymerizing , bundling , and turnover rates for the actin cytoskeleton . Our results are consistent with the observations that a minimum level of RhoA is necessary; however , hyperactivity results in destabilization and progressive loss of FPs [16] . We propose that upon inhibition of Rac1 , the cell may sustain its FP integrity with sufficient levels of actin bundles . Signaling through nephrin stimulates actin polymerization , which could also influence the parameter αf; similarly , mutations in crosslinking proteins would impact the parameter αb . Changes in either of these positive feedback parameters for polymerization may give rise to a second equilibrium point for the system , decreasing its robustness . We hypothesize that healthy cells present a set of parameters that result in a single equilibrium point , representing FPs with strong bundles . In such circumstances , all FPs are stable and have identical properties . Using a reconstructed geometry , we studied the impact of parametric inhomogeneity . Simulations using a representative geometry with features that are important for the diffusion process ( such as distances , cross sectional areas and surface area-to-volume ratios for FPs , primary processes and cell body ) were used to study the crosstalk between G-actin whole-cell diffusion and the localized polymerization of F-actin in FPs , followed by crosslinking into bundles . As demonstrated here , G-actin availability was a limiting factor , and strong localized polymerization may disrupt the cytoskeleton of FPs elsewhere . The diffusion limited G-actin availability may also disrupt otherwise synchronized oscillations . This results in slow and progressive loss of bundles , the surrogate for effacement of FPs . Of course , the sudden localized collapse of actin structures may be driven by a local inhibitory effect ( either decreased polymerization rate constant or enhanced turnover ) . However , in this work , we demonstrate the feasibility of an alternative hypothesis: enhancing polymerization in remote regions of the podocyte , by a number of signaling pathways , may sufficiently disrupt the balance of G-actin availability to irreversibly drive the local collapse of existing FPs .
Methods for processing of podocyte images and analysis of podocyte morphology are detailed in Supporting Material . The analysis of the ODE model was performed in Mathematica ( Wolfram ) and the spatial simulations of PDEs in Virtual Cell ( VCell . org ) . The full VCell model is named “Falkenberg:PodocyteStability” and may be accessed through vcell . org . Segmented and reconstructed SBEM imaging data can be obtained through Dryad data repository at doi: 10 . 5061/dryad . 09d0k . Further details of computational methods and analyses are presented in Supporting Material . | Podocytes are specialized kidney cells with intricate geometries . Regulation of cell morphology and cell-cell interactions is crucial for podocyte function within the delicate glomerular filtration unit . While the cellular cytoskeleton is dynamic , it must be tightly regulated so that the cell shape and the glomerular filter are maintained . However , genetic defects , disease or drugs may perturb the balance of the cytoskeleton regulators or produce local pressure fluctuations causing proteinuria or focal morphological defects . We derived podocyte morphology from 3-D electron microscopy and constructed a mathematical model to study the actin cytoskeleton . The model reveals the potential for local chaotic cytoskeletal instability when cytoskeleton regulators ( Rac1 and RhoA ) are imbalanced . We also explore compensatory mechanisms for pathophysiological imbalances in molecular activities . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"cell",
"motility",
"medicine",
"and",
"health",
"sciences",
"actin",
"filaments",
"cell",
"processes",
"respiratory",
"system",
"kidneys",
"cellular",
"structures",
"and",
"organelles",
"cytoskeleton",
"dynamic",
"actin",
"filaments",
"contractile",
"proteins",
"polymer",
"chemistry",
"actins",
"actin",
"polymerization",
"proteins",
"chemistry",
"polymerization",
"biochemistry",
"cytoskeletal",
"proteins",
"biochemical",
"simulations",
"thoracic",
"diaphragm",
"cell",
"biology",
"anatomy",
"biology",
"and",
"life",
"sciences",
"chemical",
"reactions",
"physical",
"sciences",
"computational",
"biology",
"renal",
"system"
] | 2017 | Fragility of foot process morphology in kidney podocytes arises from chaotic spatial propagation of cytoskeletal instability |
The highly conserved , Nxf/Nxt ( TAP/p15 ) RNA nuclear export pathway is important for export of most mRNAs from the nucleus , by interacting with mRNAs and promoting their passage through nuclear pores . Nxt1 is essential for viability; using a partial loss of function allele , we reveal a role for this gene in tissue specific transcription . We show that many Drosophila melanogaster testis-specific mRNAs require Nxt1 for their accumulation . The transcripts that require Nxt1 also depend on a testis-specific transcription complex , tMAC . We show that loss of Nxt1 leads to reduced transcription of tMAC targets . A reporter transcript from a tMAC-dependent promoter is under-expressed in Nxt1 mutants , however the same transcript accumulates in mutants if driven by a tMAC-independent promoter . Thus , in Drosophila primary spermatocytes , the transcription factor used to activate expression of a transcript , rather than the RNA sequence itself or the core transcription machinery , determines whether this expression requires Nxt1 . We additionally find that transcripts from intron-less genes are more sensitive to loss of Nxt1 function than those from intron-containing genes and propose a mechanism in which transcript processing feeds back to increase activity of a tissue specific transcription complex .
Export of mRNAs from the nucleus is a fundamental aspect of gene expression regulation . This involves RNA processing and packaging followed by translocation via nuclear pores ( reviewed in [1] , [2] ) . During transcription , nascent RNA associates with hnRNPs and a succession of export factors . The THO complex binds transcripts as they emerge from the RNA polymerase , and promotes transcription elongation and 3′ end processing [3] . Associated with THO are UAP56 ( an RNA helicase ) and REF ( an adaptor protein ) , to make a super-complex , TREX [4] . Splicing occurs co-transcriptionally , and intron excision leads to deposition of the exon junction complex ( EJC ) 5′ of the splice site [5] . The EJC is also implicated in recruitment of TREX to the nascent transcript [6] . After this initial processing , a dimer comprising Nxf1 ( TAP ) and Nxt1 ( p15 ) is recruited via interaction of Nxf1 with REF [7] . Recruitment of these export factors renders the mature poly adenylated mRNA competent for export via nuclear pores . mRNAs derived from intronless genes also interact with REF and UAP56 , and this again recruits Nxf1/Nxt1 [8] . The Nxf1/Nxt1 dimer is released from the mRNA in the cytoplasm , and recycled to the nucleus . Most RNA polymerase II transcripts are exported via this Nxf1/Nxt1-dependent route , however some endogenous mRNAs and viral RNAs use a parallel route , dependent on the nuclear protein export factor Crm1 , acting via unknown RNA-binding adapters [9] . The components of the canonical pathway are well conserved throughout eukaryotes . REF ( RNA export factor ) is also known as ALY or Aly ( ally of AML1 and LEF1 ) , however the protein product of the Drosophila melanogaster meiotic arrest transcriptional regulator aly ( always early ) is also abbreviated Aly . To avoid confusion , throughout this paper we will use REF to denote the RNA export factor , encoded in Drosophila by Ref1 and ref2 , and aly ( Aly ) to refer to the always early gene ( and protein ) . Differential gene expression underlies the morphological diversity of differentiated cell types , and depends on regulation at both transcriptional and post-transcriptional levels . The production of mature sperm demonstrates the extreme morphological specialisation achievable , and relies on testis-specific expression of many genes . About 8% of annotated Drosophila protein-coding genes are expressed exclusively in testes [10] , [11] . Transcription of these genes occurs predominantly in primary spermatocytes [12] , [13] . The mRNAs are exported from the nucleus and either translated immediately , or , more typically , are stored in translationally inactive RNPs for translation during spermiogenesis ( reviewed in [14] ) . Most of the “meiotic arrest” genes promote the spermatocyte-specific transcriptional programme ( reviewed in [15] ) . Meiotic arrest testes contain only stages up to and including mature primary spermatocytes and the spermatocytes of almost all meiotic arrest mutants fail to express a very large subset of the testis-specific transcripts , as well as some ubiquitous transcripts [16] , [17] . Most meiotic arrest genes encode proteins predicted to form two distinct complexes , tMAC and a testis-specific version of TFIID ( reviewed in [15] ) . These complexes are believed to primarily regulate transcription within primary spermatocytes , rather than acting post-transcriptionally . tMAC comprises at least 6 subunits that co-purified from testis extract , encoded by aly , comr , tomb , topi , mip40 and Caf1 [18] . Genetic and protein interaction data indicate that the complex also probably contains Wuc and Achi/Vis [16] , [19] . At least four tMAC subunits ( Comr , Tomb , Topi and Achi+Vis ) contain predicted DNA-binding motifs , and the complex localises to chromatin in primary spermatocytes , consistent with a transcriptional transactivator role [20] , [21] , [22] , [23] , [24] . The testis-specific TATA-binding proteins ( tTAFs ) , subunits of a putative testis form of TFIID , and encoded by can , mia , nht , rye and sa , localise to promoters of target genes and the nucleolus in primary spermatocytes [25] , [26] , [27] . The meiotic arrest mutants are classified according to phenotype . aly-class mutants ( aly , comr , tomb , topi and achi+vis ) have virtually undetectable levels of a very large number of transcripts in mutant testes . All aly-class mutants encode tMAC subunits . can-class mutants ( encoded by tTAF genes ) have dramatically reduced expression of a large subset of the genes affected in aly-class mutants , although they express normal levels of some aly-class dependent transcripts [15] . Interestingly , wuc;aly double mutant testes , despite being mutant for two tMAC subunits actually has a can-class phenotype with respect to defects in gene expression [16] . Recently a new meiotic arrest mutant , thoc5 , which encodes a THO complex subunit was reported [28] . No transcriptional defects were reported in this mutant , although transcription of only a few genes was assayed . Here we identify Nxt1 as a meiotic arrest gene in Drosophila , and show that it is a founding member of a new meiotic arrest gene class . Partial loss of function of Nxt1 results in failure of primary spermatocytes to accumulate many transcripts . Transcripts that were sensitive to loss of Nxt1 were also dependent on tMAC for their transcription , indicating a mechanistic link between the testis-specific transcription machinery and the core mRNA export pathway . We demonstrate that the link depends on the promoter used to generate the mRNA , and not on the mRNA sequence itself . We found that spliced transcripts are expressed more efficiently in the Nxt1 loss of function background than unspliced transcripts , and propose a model in which RNA processing feeds back to increase the activity of the tissue-specific transcription complex .
The line z2-0488 was identified by screening a collection of EMS-induced male sterile mutants using phase-contrast microscopy of live testis squash preparations [29] . z2-0488 homozygote and hemizygote males showed a typical meiotic arrest phenotype; the mutant testes had stages up to and including mature primary spermatocytes , however the meiotic divisions and spermatid differentiation were absent ( Figure 1C ) . Positional cloning ( see Materials and Methods and Figure S1 ) revealed that z2-0488 causes an amino acid substitution ( D126N ) in the RNA export protein Nxt1 . An independent allele . Nxt1DG05102 , which has a P-element insertion in the coding region , and is expected to be a null , was obtained from Bloomington Drosophila stock centre . Nxt1z2-0488 homozygotes , Nxt1z2-0488 hemizygotes and Nxt1z2-0488/Nxt1DG05102 transheterozygotes were semi-lethal and male and female sterile . Most of these pupae had head eversion defects during the pupal stages ( Figure 1 H–K ) , accounting for the semi-lethality . Testes from Nxt1z2-0488/Nxt1DG05102 displayed a meiotic arrest phenotype ( Figure 1B ) . Nxt1DG05102 homozygotes and hemizygotes were embryonic lethal . Depletion of Nxt1 from spermatocytes by RNAi also generated a meiotic arrest phenotype ( Figure 2A ) , indicating that Nxt1 functions autonomously in these cells . We conducted a detailed phenotypic analysis to compare the meiotic arrest caused by Nxt1 mutation with previously described meiotic arrest loci . RNA in situ hybridisation revealed that Nxt1 mutant testes had defects in accumulation of some tMAC target mRNAs , while other tMAC target transcripts were detected in the cytoplasm of the mutant primary spermatocytes ( Figure 3 ) . Nxt1 mutant testes failed to express some genes that were expressed in nht , a can-class mutant ( Figure 3 ) . The RNA in situ hybridisation profiles thus did not match those seen in any characterised class of meiotic arrest mutant ( Figure 3 ) , suggesting that Nxt1 is the founding member of a novel class . Surprisingly , given the known role of Nxt1 in mRNA export from the nucleus [30] , we detected no accumulation of transcripts within the nucleoplasm . All mRNAs detected were exclusively cytoplasmic ( Figure 3H–H′″ ) . Uniquely among meiotic arrest mutants , thoc5 primary spermatocytes showed severe fragmentation and disorganisation of the nucleolus [28] . Phase contrast microscopy revealed no overt nucleolar organisation defect in Nxt1 mutant testes , although the nucleoli in some mature primary spermatocytes appeared smaller than in wild type ( Figure 1D–G ) . Sa-GFP , whose localisation is abnormal in both aly-class and thoc5 mutant testes , was localised to the nucleolus and indistinguishable from wild type in Nxt1 primary spermatocyte ( data not shown ) . The chromatin in arrested spermatocytes was partially condensed , and displaced from the nuclear envelope ( Figure 1F , G ) . We determined the sub-cellular localisation of Nxt1 protein by expressing eGFP-Nxt1 using the GAL4-UAS system . When expressed in spermatocytes ( with bam-Gal4VP16 ) , the protein was primarily nuclear , although some cytoplasmic signal was detected ( Figure 4A , B ) . Within early primary spermatocyte nuclei the signal was weaker in the nucleolus , and enriched in a peri-nucleolar dot ( Figure 4E , F ) . As primary spermatocytes matured this peri-nucleolar dot disappeared , and one or more cytoplasmic puncta appeared , typically apposed to the nuclear membrane ( Figure 4G , H ) . The protein remained stable into spermatid differentiation , and was localised to one face of the nuclear envelope of early elongation stage spermatids ( Figure 4I , J ) . The protein remained detectable in the needle-shaped nuclei of spermatids ( Figure 4K , L ) , but was not detected in mature sperm . When eGFP-Nxt1 was expressed in ovarian somatic cells using a ubiquitous driver ( tubulin-Gal4 ) it was predominantly nuclear; relatively uniformly distributed within the nucleoplasm , but weaker in the nucleolus . Some cytoplasmic signal was detected ( Figure 4M , N ) . The crystal structure of the human Nxf1/Nxt1 dimer has been solved , and we used this structure to predict the molecular effects of the D126N mutation [31] . The residue is conserved in metazoa ( D131 in human Nxt1 ) , and lies on a beta sheet at the Nxt1/Nxf1 dimer interface . The D131 side chain is oriented towards the core of the protein , and participates in a hydrogen bonding network with three other residues , Y24 , Y39 and Q111 . All three of these residues are conserved throughout metazoa , suggesting that this H-bonding network is conserved . Substitution of D131N disrupts the H-bonding network ( Figure 5 ) . This is likely to destabilise the folding around the core of the protein , particularly the interaction of the beta sheets of the dimer interface and an alpha helix on the opposite protein face . Expression of GFP-Nxt1-D126N revealed a dramatic decrease in protein stability ( Figure 4 ) . When driven with bam-Gal4VP16 , fusion protein was only detected in the cells in which the driver itself was active ( Figure 4C , D ) . The fusion protein in these cells was uniformly distributed . When expressed using tubulin-Gal4 no fusion protein could be detected under conditions in which the wild type protein was detected ( Figure 4O , P ) . The mutation is not predicted to affect the dimer interaction with Nxf1 , thus we conclude that the primary defect in the mutant is reduced protein stability , but expect that any mutant protein that does get folded is likely to be able to participate in mRNA nuclear export . We used microarrays to characterise the gene expression defects in Nxt1 mutant testes . We found that Nxt1 mutant testes had dramatic defects in transcript accumulation , with many transcripts being reduced in abundance , or undetectable ( Figure 6A ) . As with other meiotic arrest loci , there were very few transcripts whose level was elevated in mutant testes compared to controls ( Figure 6A , Figure S2 ) . 485 probes showed a 16 fold or more decrease in signal in Nxt1 mutant testes compared to normal testes . Comparison with FlyAtlas data revealed that the transcripts highly dependent on Nxt1 for expression in testes were strongly biased towards testis-specific expression ( Figure 6B ) . 78% of all highly Nxt1-dependent transcripts were testis-specific ( Figure 6D ) , although clearly the majority of testis-specific transcripts are not highly Nxt1-dependent ( defined as 16× or more down-regulated in Nxt1 testes ) . To test whether the transcription defects were a peculiarity of our EMS allele , we assayed gene expression changes in the meiotic arrest testes generated by Nxt1 RNAi in spermatocytes . We found a very dramatic down-regulation of ran-like , and a mild down-regulation of expression of djl and CG42355 , mirroring the results seen with Nxt1z2-0488/Nxt1DG05102 ( Figure 2B ) . We compared the global effects on gene expression of Nxt1 mutation with microarray data from aly testes [16] . 1064 probes were highly ( 16× or more ) down-regulated in aly testes . Comparison of the fold-change in Nxt1 vs . control and aly vs . control revealed that 424 ( 87% ) of the 485 probes that were highly dependent on Nxt1 for expression were also highly dependent on aly for expression . ( Figure 6C , D; Figure S2 ) . Of the remaining 61 highly Nxt1-dependent probes , 34 were 8–16 fold down-regulated in aly , 24 were 2–8 fold down-regulated in aly and only 3 were normal or nearly-normal ( less than 2× down ) in aly . In contrast , the 1064 highly aly-dependent probes showed a graded response to loss of Nxt1 . 424 ( 40% ) were also highly dependent on Nxt1 ( 16 fold change ) , 186 ( 17 . 5% ) showed a strong requirement for Nxt1 ( 8–16 fold change ) , 313 ( 29 . 5% ) showed a mild-moderate reduction in expression ( 2–8 fold change ) while 141 ( 13% ) were detected at normal , or near-normal ( less than 2 fold change ) levels in mutant testes ( Figure 6C , D; Figure S2 ) . We did not see any reduction in the levels of transcript for any of the known tMAC subunits in Nxt1 testes ( Figure S2 ) . Immunohistochemistry indicated that Aly protein is expressed and localised to the nucleus as per wild type in Nxt1 testes ( data not shown ) . The finding that many highly aly-dependent transcripts are expressed at near-normal levels in Nxt1 testes , coupled with the observation that no known tMAC components are under-expressed , indicate that the transcript accumulation defects seen in Nxt1 testes are very unlikely to be caused indirectly , via reduction in expression of a tMAC subunit . These data confirm that Nxt1 founds a novel meiotic arrest class of mutant , and is required for accumulation in testes of a large set of predominantly testis-specific transcripts that depend on tMAC for transcription . In light of the known role of Nxt1 in RNA nuclear export we tested whether the gene expression defects in Nxt1 testes are the consequence of reduced transcription , or whether normal transcription occurs but the transcripts have reduced stability , or both . We compared the levels of nascent RNA to mRNA in wild type and mutant cells . If transcripts are produced in the mutant , and then degraded , the level of nascent RNA would be similar to WT , while the mRNA level is reduced . If transcription itself is reduced in the mutant then both RNA species would be lower than in WT . To distinguish between nascent and mature RNA we generated cDNA from DNase-treated total spermatocyte RNA using random primers . We amplified nascent RNA with at least one primer in an intron; for mRNA we placed one primer in an exon and the other spanning an exon-exon junction ( Figure 7A ) . All genes selected for this analysis had a single , small , intron . Control genes , ocn and CG12699 were selected that showed no defect in expression in Nxt1 according to the microarrays . mRNAs and nascent RNAs of these control genes were expressed in mutant spermatocytes at levels similar to WT ( Figure 7B ) . 9 genes ( CG4907 , CG10478 , CG11249 , CG14546 , CG16736 , CG17380 , CG32487 , CG33125 and pif2 ) were selected as being highly dependent on Nxt1 from the microarrays . For all these test genes we saw a reduction in the mature transcript level of 20–40 fold . Analysis of the nascent transcripts revealed a reduction of about 5 fold ( Figure 7B ) . Thus the failure of Nxt1 mutant testes to accumulate testis-specific transcripts can partially be explained by reduced transcription of these genes . The finding that Nxt1-dependent transcripts all rely on tMAC for their expression suggests that the transcriptional regulation could be key to determining whether a specific transcript requires Nxt1 for accumulation . We tested this using reporter constructs in which a testis-specific transcriptional control element is used to drive expression of LacZ ( Figure 8A ) . djl is highly transcribed in primary spermatocytes , and the transcript remains stable until late elongation , when it is translated . djl is testis-specifically expressed and depends strongly on the tMAC component aly , but less strongly on Nxt1 , for its expression ( Figure 3 , E–E′″ , Figure 8C ) . A genomic fragment spanning the region from −555 to +95 ( relative to the transcription start site , TSS ) is sufficient to drive transcription of the reporter in primary spermatocytes , and to confer translational repression on the mRNA [32] . This transcription has already been shown to be reduced in tTAF mutants [32] . RNA in situ hybridisation ( data not shown ) and q-RT-PCR ( Figure 8C ) revealed , as expected , that this djl+95-LacZ reporter construct expression is extremely low in a tMAC mutant ( comr ) . We found a dramatic reduction in reporter expression in Nxt1 mutant primary spermatocytes compared to wild type by q-RT-PCR and RNA in situ hybridisation ( Figure 8B , C ) . This indicates that the promoter and/or 5′UTR are critical for determining Nxt1 dependence . To test whether the Nxt1-dependence of a specific transcript depends on the transcript sequence or on the promoter used to activate transcription , we drove expression of the djl+95-LacZ transcript independently of the djl promoter by replacing the djl promoter with 5×UAS and the Hsp70 minimal promoter ( up to the TSS , Figure 8A ) . q-RT-PCR of spermatocyte samples revealed that reporter expression in Nxt1 mutant cells ( w ; Nxt1z2-0488/Nxt1DG05102 ; UAS-djl-LacZ/bamGal4VP16 ) was essentially equal to that in control cells ( w ; Sco/Nxt1DG05102 ; UAS-djl-LacZ/bamGal4VP16 ) when driven using bamGal4VP16 ( Figure 8C ) . The transcripts from djl+95-LacZ and UAS-djl-LacZ have the same sequence; they differ only in the promoter sequence used to activate their expression . Their differential requirement for Nxt1 in accumulation of the transcript therefore reveals that the promoter sequence , rather then the RNA sequence , imposes Nxt1-dependence on the transcript . To test whether this finding is a general property of tMAC-dependent promoters we generated a reporter using the CG42355 promoter and 5′ UTR . CG42355 is expressed specifically in testes in the same pattern as djl . CG42355 expression depends on tMAC but less strongly on Nxt1 ( Figure 3 F–F′″ , Figure 8C ) . CG42355 promoter + 5′UTR ( −177/+136 ) - LacZ reporter lines recapitulated the native expression of the mRNA in control testes ( Figure 8B ) , and revealed that CG42355 , like djl , is significantly translationally repressed ( Figure S3 ) . Expression of this construct was barely detected in both comr and Nxt1 mutant spermatocytes ( Figure 8C , Figure 9 ) . We conclude that the tMAC-dependence of a promoter + 5′UTR is sufficient to confer Nxt1-dependence onto an exogenous transcript . The native CG42355 gene must contain features that allow the transcript to overcome this requirement for Nxt1 , and be expressed at only 5× reduced levels in Nxt1 mutant cells . tMAC predominantly regulates genes whose transcripts are normally translationally repressed in the male germline; if Nxt1 is required to recognise and protect translationally repressed mRNAs then the overlap in target genes would be merely co-incidental . To test this we used a djl+43-LacZ reporter construct ( Figure 8A ) , which is transcribed in primary spermatocytes , but not translationally repressed ( Figure S3 ) [32] . We confirmed by both RNA in situ hybridisation ( data not shown ) and q-RT-PCR ( Figure 8C ) that transcription of this reporter depends on tMAC function . Expression of this reporter in Nxt1 mutant testes was even lower than expression of the djl+95-LacZ construct , indicating that the promoter and first 43 bp of 5′UTR , and not the translational repression sequence of the mRNA , is responsible for conferring Nxt1 dependence on this exogenous reporter transcript ( Figure 8B , C ) . This correlates with the results obtained from the UAS-djl+95-LacZ construct , in which an mRNA that is translationally repressed accumulates normally in Nxt1 mutant testes ( Figure 8B , Figure S3 ) . To understand why some tMAC-dependent transcripts require Nxt1 for their expression , while others show much lower dependence on Nxt1 , we further analysed the microarrays , and focused on genes 16× or more down-regulated in Nxt1 or aly mutants . Highly aly-dependent transcripts were significantly longer ( mean 1694 bp , SD 1209 ) than highly Nxt1-dependent transcripts ( mean 1518 bp , SD 1170 ) ( 2-tailed Mann-Whitney U test , p = 0 . 0024 ) . This difference was even more dramatic when the length of highly Nxt1-dependent transcripts was compared to those that depend highly on aly but are not 16× or more down-regulated in Nxt1 ( mean 1800 bp , SD 1238 ) ( 2-tailed Mann-Whitney U test , p<0 . 0001 ) . Thus we conclude that shorter transcripts are more likely to be dramatically under-expressed in Nxt1 testes than longer transcripts . However CG42355 encodes a short transcript ( <1 kb ) that is only mildly down-regulated in Nxt1 mutant testes , while the LacZ reporter driven by this promoter is much longer ( >3 kb ) , and is dramatically under-expressed in the mutant , so other features of the transcript must also be implicated . Approximately 22% of all annotated Drosophila genes lack introns . 1412 genes are expressed specifically in testis ( based on FlyAtlas data ) and 521 ( 37% ) of these lack introns , a significantly higher proportion than expected by chance ( Chi-squared , p<0 . 0001 ) . Thus , intron-less genes are enriched for testis-specific expression . Of the 1019 genes that were 16× or more reduced in aly testes , 404 ( 40% ) lack introns , indicating that the presence or absence of introns does not affect whether a testis-specific transcript is also highly aly-dependent ( Chi-squared , p = 0 . 18 ) . In contrast , of the 456 highly Nxt1-dependent genes , 303 ( 66% ) lack introns , revealing that highly Nxt1-dependent genes are significantly more likely to lack introns than randomly selected testis-specific genes ( Chi-squared , p<0 . 0001 ) . Genes that are highly dependent on both aly and Nxt1 for their expression are predominantly intron-less while genes that are highly dependent on aly , but are only mildly dependent on Nxt1 predominantly have introns ( Figure 6D ) . Notably , the more introns present in a strongly tMAC-dependent gene , the higher its expression in an Nxt1 background ( Figure S4 ) . We analysed the 300 most down-regulated genes in a tMAC mutant , and correlated fold change compared to control in Nxt1 mutants with intron number ( 0 , 1 , 2 , 3+ ) . Relative expression of genes lacking introns was significantly lower than that of genes containing introns ( t-test , p = 1 . 18889E-32 ) . Relative expression of genes with 1 intron was not significantly different than that of 2-intron genes ( t-test , p = 0 . 126 ) , however genes with 1 or 2 introns had significantly lower expression in Nxt1 mutants compared to wild type than genes with 3 or more introns ( t-test , p = 0 . 00013 ) . This indicates a dose-response relationship between gene expression in Nxt1 mutant testes and intron number , where the more introns a tMAC-dependent gene , has the lower its requirement for Nxt1 for expression . The microarray results indicate that having at least one intron is beneficial for expression of tMAC-dependent genes in Nxt1 mutants . We used reporter constructs to test how the presence or absence of introns impacts on reporter gene expression in Nxt1 testes ( Figure 8A ) . We inserted the djl intron into the 5′UTR of ( the previously intron-less ) djl −555/+95 reporter sequence , and found that the reporter was testis-specifically expressed , and translation was delayed until spermatid elongation stages , in a pattern indistinguishable from the original construct . Similarly , insertion of the CG42355 introns into the LacZ coding sequence of the djl construct resulted in expression mirroring the original reporter ( Figure S3 ) . Equivalent one and two intron CG42355 reporters were constructed ( Figure 8A ) . We confirmed correct splicing of these transcripts by RT-PCR ( data not shown ) . When we tested expression of these intron-containing reporters in Nxt1 spermatocytes we found that insertion of the introns partially restored the reporter expression ( Figure 8D , Figure 9 ) . Strikingly , the djl- reporter with two introns was expressed at over 50% of wild type levels . Similarly , introduction of two introns into the CG42355 reporter elevated expression to 15% of that seen for the reporter in wild type spermatocytes . This represents a 10× increase in expression in mutants compared to the intron-less version of the reporter . Insertion of introns did not increase reporter expression in comr mutant testes ( Figure 9 ) . We conclude that , in spermatocytes , tMAC-dependent promoters confer dependence on Nxt1 for transcript expression , but that this dependence can be partially overcome if the transcript has one or more introns .
To date 14 meiotic arrest loci have been described ( aly , achi-vis , topi , tomb , comr , wuc , mip40 , can , mia , sa , rye , nht , thoc5 , Nurf301 ) [15] , [28] , [33] . Most of these encode subunits of either a testis-specific TFIID complex or tMAC . The exceptions are mutations resulting in a C-terminally truncated product from Nurf301 [33] , and loss of thoc5 [28] . In both these cases the effect on gene expression is apparently much less dramatic than loss of either tTAFs or tMAC . Here we identify Nxt1 as a novel meiotic arrest locus , and show that the phenotype does not fit with any previous classification . Notably , mutation of Nxt1 has much more dramatic effects on gene expression than thoc5 , Nurf301 or tTAFs , although the defect is not as pronounced as that seen in most tMAC mutants . Nxt1 protein acts in the same biochemical pathway as the TREX subunit , Thoc5 , suggesting that the entire RNA export pathway might be critical for testis gene expression . In support of this , in preliminary experiments we have observed incomplete meiotic arrest phenotypes in testes with spermatocyte specific knock down ( by RNAi ) of either Ref1 or sbr ( data not shown ) . No defects in gene expression were detected in the thoc5 mutant [28] , but we note that the genes assayed ( bol , twe , cycB , dj , fzo , Mst87F ) all contain at least one intron , and , although fzo , bol and dj have reduced expression in Nxt1 testes , none is on the list of genes 16× or more down regulated in this background . It would be very interesting to determine whether the intron-less genes most dependent on Nxt1 for expression are also dependent on thoc5 or other RNA export factors . Notably , Nxt1 and Thoc5 localise to a punctate structure adjacent to the nucleolus in early primary spermatocytes and to a structure adjacent to the nuclear envelope in late primary spematocytes . This localisation adjacent to the nuclear envelope is also seen for Nxf1 , the binding partner of Nxt1 , encoded in Drosophila by the sbr gene [28] , [30] , [34] . Both Nxt1 and Sbr also localise to one face of the nucleus in spermatids . This co-localisation suggests that the proteins are all working together in these cells . The direct link we describe here , between tMAC and RNA export in regulating testis gene expression , can be explained by two theoretical mechanisms . ( 1 ) tMAC could promote transcription of target genes and feed-forward to promote their stability and RNA export . In the absence of Nxt1 the mRNAs would not be protected , and instead would be degraded . ( 2 ) Initially low tMAC activity promotes production of low levels of transcript . Binding of the export factors to these transcripts during processing would feed-back to increase the activity of tMAC . This would then increase the transcription of the tMAC-dependent gene . The ability of introns to rescue expression in the mutant indicates that the mechanism must be mediated through the RNA transcript , and presumably depends on the higher affinity of spliced transcripts for the Nxt1/Nxf1 dimer . If the feed-forward mechanism predominates , the rate of nascent transcript production should be equal to wild type . If feed-back predominates , the rate of nascent transcript production would be lower than wild type . We found a reduction in both nascent and mature transcripts in the mutant , although the reduction in mRNA is greater than the reduction in nascent RNA . We propose that both feed-forward and feedback occur , to give an amplification loop . This amplification would depend on a specific interaction between the RNA export factors and tMAC . The RNA nuclear export pathway has been deduced primarily using experiments in yeast , and in tissue culture cells ( reviewed in [1] ) . Drosophila culture cells depleted of components of the pathway , including Nxt1 , accumulate poly-adenylated mRNAs within the nucleus . Most transcripts in the cell are also reduced in abundance , typically a 1 . 5–2 fold level change for altered transcripts [35] . Very few transcripts were reduced more than 5 fold after RNAi treatment , and very few were significantly increased in abundance . The RNAi treatments used in these experiments also caused defects in cell proliferation or viability . The in vivo role of sbr , which encodes the Nxf1 partner of Nxt1 , has been investigated in embryos and larvae , and again defects have been detected in bulk export of mRNAs from nuclei [36] . In none of these experiments has a link been detected between the role of the Nxt1/Nxf1 and transcription . Factors in the mRNA export pathway , upstream of Nxt1/Nxf1 have been linked to transcription , consistent with the fact that RNA splicing and processing is co-transcriptional . Specifically the TREX complex interacts with chromatin and RNA polymerase II , and is important for facilitating transcription elongation , 3′ end formation , mRNA processing , and transfer to the nuclear pore [4] . While the interaction of TREX with RNA polymerase is well established , there has been no data to implicate Nxt1/Nxf1 in this process . TREX promotes association of REF with the transcript , and this in turn promotes Nxt1/Nxf1 association [37] . Similarly , the spliceosome can have a stimulatory effect on transcription , both at the level of initiation and elongation [38] . Our study is the first , to our knowledge , to implicate these pathways in the regulation of a programme of gene expression promoted by a specific transcription factor complex . Recently two studies have demonstrated that the stability of specific S . cerevisiae mRNAs can depend on non-transcribed promoter sequences . RPL30 mRNA was shown to have a short half-life imposed by the transcriptional activator Rap1 and its binding site in the UAS , although how this affects the RNA is not yet determined [39] . The cell cycle regulation of the half-lives of SWI5 and CLB2 was independent of the transcript sequences , was coordinated with transcription , and was promoter dependent [40] . Dbf2 , a kinase , is recruited by specific promoter sequences and co-transcriptionally deposited on the SWI5 and CLB2 mRNAs . Dbf2 is a component of the CCR4-NOT cytoplasmic deadenylase complex , and the activity of this complex could then normally de-stabilise these transcripts . In both these examples the promoter acts to confer a short half-life on the transcripts it regulates . We report a similar phenomenon in an animal system , however the outcome is increased mRNA expression , with the RNA processing pathway feeding back to increase transcription . The canonical role for Nxt1 is export of mRNAs from the nucleus , however , paradoxically , and in contrast to the results from tissue culture studies [41] , we do not detect accumulation of mRNAs in Nxt1 mutant nuclei . The majority of mRNAs expressed in the mutant spermatocytes are present at normal levels , and are detected in the cytoplasm . This mRNA export capability could be provided by residual Nxt1 function , or by a parallel pathway . We suggest instead that residual Nxt1 function provided by the hypomorphic allele is sufficient for mRNA export . This interpretation is supported by the observation that the null allele is embryonic lethal while the z2-0488 allele is viable , thus sufficient activity remains in this allele to support normal function of most cells . The fact that reduction in Nxt1 levels by RNAi in spermatocytes phenocopies the z2-0488 allele indicates that the effect is due to reduced activity , rather than an allele-specific effect . A well characterised alternative mRNA export pathway involves the Crm1 protein , binding RNA via an unidentified adaptor protein . Crm1 is implicated in export of a subset of endogenous RNAs , as well as HIV mRNA in humans [42] . The Drosophila Crm1 protein is encoded by the gene embargoed . This gene is expressed at much lower levels in testes than in other adult tissues , and thus is unlikely to represent the major RNA export pathway in spermatocytes , given their high level of transcriptional activity . Nxt1/Nxf1 heterodimers are already known to be implicated in export of some mRNAs with retained introns , and have been demonstrated to interact with constitutive transport elements present in some cellular and viral mRNA [43] . Similarly , in human cells naturally intronless transcripts tested were transported from the nucleus in a TREX and Nxf1 dependent manner [44] . In the absence of Nxf1 the transcripts remained nuclear , but apparently were not destabilised . When nascent transcripts are being produced they are bound by hnRNPs , and those with introns are processed by the spliceosome . During this processing the EJC associates with the transcript . This is then responsible for recruitment of Ref1 , which recruits Nxt1/Nxf1 . Transcripts that lack introns also associate with Ref1 and thus Nxt1/Nxf1 , but without the help of the EJC . Thus the EJC increases the affinity of Nxt1/Nxf1 to the transcript . For a transcript being synthesised in an Nxt1-depleted background this could be sufficient to ensure association of the export factor , and thus to ensure proper processing and export . We detect a dose response for intron number , specifically , the more introns a tMAC dependent gene has the higher its expression in an Nxt1 mutant background . If , as we propose , there is feedback from the export pathway to increase tMAC activity , then the presence of the EJC and the increased affinity for Nxt1/Nxf1 would result in higher transcript levels from intron-containing genes than intron-less genes in an Nxt1-depleted background , consistent with our findings . It is well demonstrated that the presence of introns in primary transcripts correlates with higher gene expression levels [45]; this is true for almost all Drosophila adult tissues . Indeed , most ectopic expression systems include at least one intron to facilitate higher transcription levels , higher transcript stability and higher translation efficiency . However the testis is unique among Drosophila adult tissues in that , amongst genes whose expression is detected , the expression level of intron-containing genes is significantly lower than that of intron-less genes ( Mann-Whitney U test , p<0 . 001 ) ( deduced from FlyAtlas data; expression is in arbitrary units ) . Specifically , the median expression signal for intron-containing genes in brain ( a typical somatic tissue ) is 139 , while that for intron-less genes is 78 . 1 . In contrast , the equivalent figures for testes are 84 and 168 . 5 . In testis , and in somatic tissues , approximately 60–65% of all annotated intron-containing protein coding genes are expressed . In somatic cells only 30–35% of all annotated intron-less genes are expressed . In contrast , in testes , 63% of all annotated intron-containing genes are expressed ( Figure S5 ) . Thus spermatocytes must have evolved a mechanism to support high level expression of intron-less genes . Many of the intron-less genes expressed in testes are retroposed copies of intron-containing genes . Typically the parent gene will have a broad or ubiquitous expression pattern while the retroposed gene's expression is highly restricted [46] . The new gene is often subject to rapid evolution under positive selection , and contributes to sperm-specific cell biology . Most of these retroposed gene copies are regulated by the tMAC transcriptional regulation complex , and , like other tMAC-dependent genes , they typically have short promoters . Surprisingly , the promoter regions , and even transcription start sites , are not highly conserved between species [47] , although the insertions are biased towards genomic regions already containing testis-specific transcription units [48] . This could be explained by low-affinity tMAC binding occurring by virtue of the chromosomal domain , promoting transcription of the newly inserted sequence , and feedback from RNA export factors serving to increase tMAC activity , and thus expression of the new gene . In this study we have demonstrated a link between the core RNA export pathway and testis-specific transcription . Null alleles of Nxt1 are recessive lethal , and the hypomorph ( homozygous or hemizygous has low viability . The animals often pupate with a distinctive elongated curved shape , uneverted spiracles , and then fail in head eversion , consistent with defects in air bubble movement . These processes are controlled by transcriptional responses downstream of ecdysone signalling , and the mutant phenotype is highly reminiscent of mutants in Eip74EF , an ets family transcription factor [49] . Nxt1z2-0488 homozygous or hemizygous adult females are also sterile . These highly reproducible defects potentially also arise as a result of defects in gene expression in the relevant tissues , rather than being caused by a general defect in export of all mRNAs from the nuclei of somatic or female germline cells . In testes we have demonstrated that the effect is via a specific transcription complex , tMAC . The soma and female germline defects cannot be caused by an interaction between Nxt1 and tMAC since tMAC expression is restricted to testes . We postulate that Nxt1 might be important for regulating the transcriptional response to ecdysone during pupariation , and expression of specific genes in the female germline . Thus we suggest that Nxt1 could work with , as yet unidentified , specific transcription factors to control gene expression in other tissues .
We mapped the male and female sterility , and semi-lethality , of z2-0488 using deficiency chromosomes to a region of approximately 100 kb containing 22 known or predicted genes at the distal end of chromosome 2R ( Figure S1 ) . 9 candidate genes were sequenced , and a single non-synonymous change was found ( D126N , codon GAT-AAT ) in the Nxt1 gene . A P-element insertion allele of Nxt1 , P{wHy}DG05102 , failed to complement the semi-lethality and sterility of the z2-0488 mutation . This P-element insertion , which is in the first exon , just downstream of the ATG , and predicted to be a null allele , results in homozygous and hemizygous lethality . We conclude that z2-0488 is a hypomorphic allele of Nxt1 , and designate the allele Nxt1z2-0488 . Testes were dissected from young adults , prepared for phase contrast and fluorescence microscopy as in [50] and imaged using a Hamamatsu Orca-05G camera driven by HCImage software on an Olympus BX50 microscope . RNA in situ hybridisation using dig-labelled RNA probes was as in [51] . Primer sequences used to generate templates are in Text S1 . CycB probe was made from a cDNA clone as in [17] . Templates for CG42355 and djl were subcloned into pGEM-T-Easy , while other probes were generated directly from PCR products . Beta-galactosidase activity assays on whole mount testes were as in [17] . Colour images were taken with a JVC-F75U camera run by KY-LINK software mounted on the Olympus BX50 microscope . Image composites were assembled using Adobe Photoshop . djl-lacZ −555/+95 and djl-lacZ −555/+43 [32] were provided by Renate Renkawitz-Pohl . PCR from genomic DNA yielded a fragment 5′-EcoRI-CG42355-177/+136-BamHI-3′ which was cloned into pCaSpeR4-AUG-betaGAL [52] to yield CG42355-LacZ −177/+136 . The plasmids djl-lacZ -555/+95 +1 intron , djl-lacZ −555/+95 +2 introns , djl-lacZ −555/+43 +1 intron , djl-lacZ −555/+43 +2 introns , CG42355-lacZ −177/+136 +1 intron and CG42355-lacZ +2 introns were derived from these basic promoter-reporter constructs by incorporation of a synthetic DNA fragment containing the intron ( s ) ( see Text S1 for sequences and construction details ) . pUAST-eGFP-Nxt1 and pUAST-eGFP-Nxt1-D126N were generated by PCR and subcloning of the Nxt1 ORF from Nxt1z2-0488/CyO flies [53] . pCaSpeR4-UAS-djl- AUG-betaGAL was made by cloning a synthetic fragment comprising 5× UAS; Hsp70 minimal promoter ( up to the TSS ) ; +1–+95 of djl into the MCS of pCaSpeR4-AUG-betaGAL using EcoRI and BamH1 ( Text S1 ) . Drosophila were maintained on cornmeal , sucrose , yeast , agar medium , at 25°C ( 29°C for RNAi crosses ) . Mutant alleles were aly5 , comrZ1340 , nhtZ2-5946 , Nxt1DG05102 , Nxt1z2-0488 . w1118 was used as a control . For RNAi against Nxt1 we used UAS-hairpin lines from the Vienna Drosophila Resource Centre [54] driven by bam-Gal4VP16 , with both constructs heterozygous . bam-Gal4VP16 was also used to drive fusion protein constructs in spermatocytes . tub-Gal4 was from the Bloomington Drosophila stock centre and was used to drive ubiquitous fusion protein expression . Transgenic lines were selected after injection of P-element constructs into w1118 using standard methods . Insertions on the third chromosome were selected , balanced , and crossed into Nxt1z2-0488 , Nxt1DG05102 and comrZ1340 mutant backgrounds . For q-RT-PCR , total RNA was extracted from purified spermatocyte samples using the RNAqueous micro kit ( Invitrogen ) . Testes were dissected from 1–3 adult males of the appropriate genotype in TB ( 183 mM KCl , 47 mM NaCl , 10 mM Tris pH7 . 4 ) , transferred to a small drop of TB on a hydrophobic plate , and the testis tip was cut open using a tungsten needle . Spermatocytes were pushed out of the testis sheath into the buffer with the needle and the remainder of the testis discarded . Most of the buffer was removed , leaving approximately 1–2 µl , and 10 µl lysis solution was added . The lysate was added to 90 µl lysis solution in a microfuge tube; immediately frozen in liquid nitrogen , and stored at −80°C . Preparation of each sample took no more than 10 minutes . RNA was purified according the manufacturer's instructions . RNA was eluted from the spin column with 2×6 . 5 µl of elution buffer . The total eluate volume was approximately 11 µl . For conventional q-RT-PCR , cDNA was synthesised using Superscript III ( Invitrogen ) and oligo-dT primers . For analysis of nascent RNA and spliced products the purified RNA was treated with DNase I , half of the RNA sample was reserved as a control to assay DNA contamination and the remainder was reverse transcribed using Superscript III and random hexamer primers . PCR primers ( sequences in Text S1 ) were used that recognised exon , intron or exon-exon junction sequences . The cDNA reaction was diluted to 60 µl with water , and 1 µl of this was used as a template in PCR reactions , using PowerSybr reagent ( ABI ) in a Chromo4 instrument ( MJR ) . For q-RT-PCR of reporter gene expression and CG42355 and djl levels , the entire RNA eluate was used for cDNA synthesis with oligo dT primers . CG3927 was used as a control gene for relative quantitation using the ΔΔCt method . Expression of the test gene relative to CG3927 gene was normalised to 1 in the control sample . Control testes had the relevant transgene present in the same copy number as the test sample , but had normal testis morphology . CG3927 expression is restricted to primary spermatocytes ( Figure 3 ) . CG3927 expression in Nxt1 and comr mutant testes is similar to wild type , as judged by both RNA in situ hybridisation and microarray analysis . All reactions were performed in triplicate . We performed biological replicates for the djl+95 and djl+43 lines , for which only one insertion was available . For other reporters we used two different third-chromosome insertion lines , and show both results . Testes from Nxt1z2-0488/Nxt1DG05102 transheterozygote males , raised at 25°C were used for Affymetrix microarray analysis , and were compared to our existing microarray data sets . Sample preparation , processing and data analysis were as described in [16] . The FlyMine interface was used to extract lists of probes associated with intron-containing and intron-less genes , as well as to extract data from FlyAtlas on gene expression profiles in adult and larval tissues . A list of 1412 genes with testis-specific expression ( 1523 probes ) was created by filtering of FlyAtlas data [11] . We selected those probes where the present call in testis was 4/4 , and sum of present calls in all other adult tissues was <4 . FlyMine ( v27 . 0 , Feb 2011 ) was used to generate lists of genes with and without introns . First a list of all annotated Drosophila melanogaster genes was created . A list of genes with introns were selected using the “Gene → Introns” template , and selecting the column “genes” as output . A list of all genes without introns was generated by subtracting the genes with introns list from the all genes list . The “Gene → Affymetrix probe” template was used to generate a list of all probes for intron-less genes . This was then combined with our Affymetrix data in Microsoft Excel to allow us to analyse intron-containing and intron-less genes separately . The number of introns present in each of the 300 genes most down regulated in tomb was determined by first ranking Affymetrix fold change data for tomb-vs . -control , then manually checking annotations in FlyBase . Annotated transcripts were cross-referenced with the RNA-seq data for adult males to ensure accuracy . All genes with valid male- ( probably testis- ) expressed spliced transcripts were scored as having an intron , even if they also encoded alternative intron-less transcripts . | In multicellular organisms , differentiated cells have a cell-type specific profile of gene expression . Sperm production is particularly specialised , and as a result over 5% of all genes are expressed exclusively in the sperm precursor cells , termed primary spermatocytes . Expression of these genes depends on a particular transcription regulation complex ( tMAC ) only active in spermatocytes . In this paper we show that a factor , Nxt1 , whose previously characterised function is in transport of RNA from the cell nucleus to the cytoplasm , is also required for expression of many testis-specific transcripts . Spermatocytes deficient for Nxt1 fail to express many tMAC-dependent genes , and we show that this effect is due in part to reduced transcription . We further show that processing of the RNA , via splicing , can partially offset the need for Nxt1 in expression of tMAC-dependent genes . Our data reveal an unexpected link between the core RNA processing pathway and a tissue-specific transcription factor . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"developmental",
"biology",
"dna",
"transcription",
"rna",
"transport",
"rna",
"processing",
"gene",
"expression",
"genetics",
"gene",
"regulation",
"molecular",
"genetics",
"gene",
"splicing",
"biology",
"evolutionary",
"biology",
"cell",
"differentiation",
"evolutionary",
"developmental",
"biology",
"gene",
"function"
] | 2013 | The RNA Export Factor, Nxt1, Is Required for Tissue Specific Transcriptional Regulation |
Short-term presynaptic plasticity designates variations of the amplitude of synaptic information transfer whereby the amount of neurotransmitter released upon presynaptic stimulation changes over seconds as a function of the neuronal firing activity . While a consensus has emerged that the resulting decrease ( depression ) and/or increase ( facilitation ) of the synapse strength are crucial to neuronal computations , their modes of expression in vivo remain unclear . Recent experimental studies have reported that glial cells , particularly astrocytes in the hippocampus , are able to modulate short-term plasticity but the mechanism of such a modulation is poorly understood . Here , we investigate the characteristics of short-term plasticity modulation by astrocytes using a biophysically realistic computational model . Mean-field analysis of the model , supported by intensive numerical simulations , unravels that astrocytes may mediate counterintuitive effects . Depending on the expressed presynaptic signaling pathways , astrocytes may globally inhibit or potentiate the synapse: the amount of released neurotransmitter in the presence of the astrocyte is transiently smaller or larger than in its absence . But this global effect usually coexists with the opposite local effect on paired pulses: with release-decreasing astrocytes most paired pulses become facilitated , namely the amount of neurotransmitter released upon spike i+1 is larger than that at spike i , while paired-pulse depression becomes prominent under release-increasing astrocytes . Moreover , we show that the frequency of astrocytic intracellular Ca2+ oscillations controls the effects of the astrocyte on short-term synaptic plasticity . Our model explains several experimental observations yet unsolved , and uncovers astrocytic gliotransmission as a possible transient switch between short-term paired-pulse depression and facilitation . This possibility has deep implications on the processing of neuronal spikes and resulting information transfer at synapses .
Activity-dependent modification of synaptic transmission critically moulds the properties of synaptic information transfer with important implications for computation performed by neuronal circuitry [1]–[4] . Multiple mechanisms could coexist in the same synapse , regulating the strength or the efficacy of synaptic transmission therein in a way that depends on the timing and frequency of prior activity at that same synaptic terminal [5] . One widely studied mechanism responsible for the dependence of synaptic transmission on past activity has been dubbed presynaptic short-term plasticity [6] . Upon repetitive action potential stimulation , the response of a presynaptic terminal – usually assessed as the amount of neurotransmitter molecules released from this latter – will not follow with uniform strength but will be modified in a time- and activity-dependent manner , leading either to facilitation or to depression of synaptic release , or to a mixture of both [2] . Such stimulus-related variations of presynaptic response can span a time scale from few milliseconds to seconds from the stimulus onset [2] , [7] and fade away after sufficiently prolonged synaptic inactivity [3] , [5] . The ability of a presynaptic terminal to convey stimulus-related information is determined by the probability to release neurotransmitter-containing vesicles upon arrival of action potentials [3] , [6] . The release probability depends on the number of vesicles that are ready to be released , i . e . the readily releasable pool , but also on the state of the calcium ( Ca2+ ) sensor for the exocytosis of synaptic vesicles [8] . On the mechanistic level , both the finite size and the slow post-stimulus recovery of the readily releasable pool , that is the reintegration of the content of synaptic vesicles , give rise to the phenomenon of short-term presynaptic depression , with the extent of depression being determined by the frequency of prior synaptic stimulation [9] . The dependence of short-term facilitation on the pattern of synaptic activation is likely determined either by the slow removal of free presynaptic residual Ca2+ or by the slow unbinding of this latter from the Ca2+ sensor [3] , although these issues are still debatable [10] , [11] . Given the important role assumed by presynaptic short-term plasticity in neural computation [6] , [12] and the variety of plastic responses – depression , facilitation or both – exhibited by central synapses [13] , [14] , it is important to unravel the mechanisms that might govern dynamical transitions between depressing and facilitating synapses . The goal of the present work was to investigate one such candidate mechanism: modulation of presynaptic plasticity by glial cells and astrocytes in particular . Recent years have witnessed mounting evidence on a possible role of glial cells in the dynamics of neuronal networks [15] . In particular , the specific association of synapses with processes of astrocytes – the main type of glial cells in the hippocampus and the cortex [16]–[18] – together with the discovery of two-way astrocyte-neuron communication [19] , [20] , suggest an active role of these cells in modulation of synaptic transmission and information processing in the brain [21] . Astrocytes could modulate synaptic transmission at nearby synapses by releasing neurotransmitter ( or “gliotransmitter” ) in a Ca2+-dependent fashion [22] . In the hippocampus in particular , several studies have shown that astrocyte-released glutamate modulates short-term plasticity at excitatory synapses either towards depression or facilitation [23]–[25] . This is achieved by activation of presynaptic glutamate receptors [26] ( see also Figure 1 for a schematic presentation ) . Thus , astrocytes are equipped with means to modulate the extent to which presynaptic terminal exhibits short-term depression or facilitation in response to sustained rhythmic stimulation [27] . We devised a biophysically plausible computational model to investigate the characteristics of astrocyte modulation of presynaptic short-term plasticity . Using the model , we were able to identify the parametric regime in which the synaptic response to action potential stimulation can switch from facilitating to depressing and vice versa . This ability to switch synaptic modus operandi depended critically on the characteristics of astrocyte-to-synapse signaling . These findings highlight the new potential role played by astrocytes in defining synaptic short-term plasticity and could explain contradicting experimental evidences . Although based on experimental results in the hippocampus , [28]–[34] , our description could also be extended to model other recognized neuron-glia signaling pathways such as GABAergic gliotransmission on interneuron-to-pyramidal cell synapses in the hippocampus [35] , glia-mediated ATP release both on hippocampal synapses [36] , [37] or in the hypothalamus [38] as well as in the retina [39] , and glial modulation of neuromuscular transmission [40]–[42] .
Regulation of synaptic transmission by astrocyte-released gliotransmitter is supported by an elaborate signaling network schematized in Figure 1 . Here , we consider the well-characterized experimental case of glutamate-mediated astrocyte regulation of synaptic transmission in the hippocampus [27] , [43] . At excitatory synapses there , astrocytes can respond to synaptically-released glutamate by intracellular Ca2+ elevations that in turn , may trigger the release of further glutamate from the astrocytes [22] , [44] . This astrocyte-released glutamate ( GA ) diffuses in the extrasynaptic space and binds to presynaptic metabotropic glutamate receptors ( mGluRs ) or NMDA receptors ( NMDARs ) on neighboring presynaptic terminals [21] , [30] . Glutamate activation of these receptors can modulate Ca2+ influx into the presynaptic terminal , affecting the release probability of glutamate-containing synaptic vesicles [26] . Thus , glutamate release from the presynaptic terminal is expected to increase the astrocytic intracellular Ca2+ , eventually leading to glutamate release from that astrocyte . In turn , astrocytic glutamate modulates presynaptic Ca2+ and thus affects the amount of glutamate released from that same synapse in response to action potentials that will follow [27] . Astrocyte Ca2+ dynamics may also not be modulated by glutamate originating from the very presynaptic terminal that is regulated by the astrocyte , but rather by an exogenous source [45] . This could correspond to the heterosynaptic case whereby two distinct synapses , A and B , are contacted by processes from the same astrocyte [21] . Glutamate released by the presynaptic terminal of synapse A modulates astrocytic Ca2+ , leading to modulation of glutamate release from the presynaptic terminal of synapse B . Alternatively astrocyte Ca2+ dynamics could be modulated by intercellular IP3 diffusion from neighboring astrocytes through gap junctions [46] or by exogenous stimulation of the astrocyte by different techniques or external stimuli [47] , [48] , or occur spontaneously [49] , [50] . Although both homosynaptic and non-homosynaptic scenarios equally occur physiologically [21] , [45] , here we focus only on the latter . This approach , which is often adopted in the majority of experiments [30]–[32] , [49] , presents several advantages . First , it allows us to characterize the effect of astrocytic glutamate on short-term synaptic plasticity in general , that is , independently of the nature of synaptic inputs . Second , it uses Ca2+ signals to merely trigger glutamate exocytosis from the astrocyte . Thus we can focus on the timing of glutamate release without considering the complexity of the underlying Ca2+ dynamics [48] which can be ultimately modeled by simple stereotypical analytical functions ( Text S1 , Section I . 2 ) . Third , it can be used in the derivation of a mean-field description of synaptic transmission [51] , [52] aimed at understanding regulation of short-term synaptic plasticity by a large variety of astrocytic glutamate signals impinging on the synapse , without the need to consider an equally large number of cases .
We first studied the effect of astrocytic glutamate release on the transfer properties of our model synaptic terminal . Because the response of a synapse to action potential critically depends on the value of U0 ( equation 3 ) , which in turn could be modulated by astrocytic glutamate binding to presynaptic glutamate receptors ( equation 5 ) , we expected that the steady-state frequency response of a synapse ( ) could also be modulated by the astrocyte-synapse signaling . Since both geometry of synaptic bouton and diffusion of glutamate in the extracellular space are beyond the scope of the present work , we implicitly assumed , based on experimental evidence [31] , that the release site of astrocytic glutamate apposes targeted presynaptic glutamate receptors . When the intracellular Ca2+ in the astrocyte crossed over the threshold of glutamate exocytosis ( Figure 4A , top , dashed red line ) , the extracellular concentration of glutamate in proximity of presynaptic receptors first increased rapidly and then decayed exponentially at rate Ωc , as a result of the concomitant uptake by astrocytic glutamate transporters and diffusion away from the site of exocytosis ( Figure 4A , middle ) ( see also Text S1 , Section I . 4; Figure S5 ) . For α = 0 , equations ( 5–6 ) predict that this glutamate peak should lead to a sharp decay of U0 , followed by a slower recovery phase ( Figure 4C , left ) . Using equation ( 5 ) , we can also predict the resulting dependence of the steady-state synaptic response on the input frequency ( Figure 4C , middle ) . In the absence of astrocytic glutamate release ( thick dashed black line ) , monotonously decreases for increasing input frequency fin for the merely depressing synapse considered in this figure . At the release of astrocytic glutamate ( Figure 4A , middle ) , the peak of bound presynaptic receptors ( Figure 4A , bottom ) and the resulting sharp drop of U0 ( Figure 4C , left , black mark ) induce a strong decrease of the steady-state amount of released resources at low to intermediate input frequencies ( 0 . 1–10 Hz ) ( Figure 4C , middle , thick red line ) . In addition , the steady-state response loses its monotonicity and displays a peak frequency characteristic of facilitating synapses ( see “Mechanisms of short-term presynaptic plasticity” in “Methods” ) . The curve then slowly transforms back to its baseline form ( thin colored lines ) and the peak synaptic input frequency appears to progressively shift toward smaller input frequencies ( thick dashed arrow ) . Hence , for α = 0 , the limiting frequency ( equation 4 ) is predicted to sharply increase following astrocytic glutamate release and then to slowly relax back to smaller values ( Figure 4C , right ) . The exact opposite picture instead describes the scenario of α = 1 ( Figure 4B ) . In this case , the parameter U0 increases upon astrocytic glutamate release ( Figure 4B , left ) causing a dramatic increase of the steady-state response for a range of frequencies within 0 . 1–10 Hz ( Figure 4B , middle ) . Accordingly , the limiting frequency of the synapse dramatically reduces following astrocytic glutamate release , and slowly recovers back to its baseline value ( Figure 4B , right ) . Taken together , the above results of the mean-field analysis predict that , depending on the parametric scenario , astrocyte can either transiently decrease , when α = 0 , or increase , if α = 1 , the release of a model synapse . To assess the validity of these predictions , we show in Figure 5 the responses of two different model synapses ( A: depressing; B: facilitating ) to Poisson spike trains delivered at frequency fin ( Figure 5 , top panels for specific realizations of such spike trains ) . To simplify the presentation , we considered the case in which a single Ca2+ peak ( Figure 5 , middle ) is sufficient to trigger the release of glutamate from the astrocyte . The synaptic response under different scenarios of astrocytic glutamate modulation ( A: α = 0; B: α = 1 ) is then compared to the “Control” scenario obtained for the model synapses without astrocyte . In the case of α = 0 ( Figure 5A , bottom ) the amount of resources released by the model synapse steeply decreased at the onset of glutamate release from the astrocyte ( green area ) and slowly , i . e . tens of seconds , recovered to the levels comparable to those of the control scenario ( blue area ) . The opposite effect was observed instead for α = 1 ( Figure 5B ) . The synaptic response in this case was strongly augmented by astrocytic glutamate ( magenta area ) and then slowly decayed back to the levels obtained in control conditions ( Figure 5B , bottom ) . Collectively our mean field analysis ( Figure 4 ) and simulations ( Figure 5 ) suggest that glutamate release by the astrocyte can induce STD or STP of synaptic response to action potentials ( Figure S6 ) . Which one between these scenarios occurs depends on the value of the “effect” parameter α that lumps together both the density and the biophysical properties of presynaptic receptors targeted by astrocytic glutamate . These results are consistent with a large body of experimental observations in the hippocampus , where astrocyte-released glutamate could transiently decrease [33] or increase the synaptic response to stimulation [30]–[32] , [34] , [49] .
The character of synaptic information transfer is shaped by several factors [2] . Synaptic strength at any given moment is determined by an earlier “activation history” of that same synapse [3] , [5] . Structural and functional organization of presynaptic bouton affects the release and reintegration of neurotransmitter vesicles , ultimately defining the filtering feature ( depressing or facilitating ) of a synapse in response to spike train stimulation [3] , [68] . Existing models of synaptic dynamics assume “fixed plasticity mode” , in which the depression/facilitation properties of a synapse do not change with time . However , in biological synapses , plasticity itself seems to be a dynamic feature; for example , the filtering characteristics of a given synapse is not fixed , but rather can be adjusted by modulation of the initial release probability of docked vesicles [13] . Using a computational modeling approach , we showed here that astrocytes have the potential to modulate the flow of synaptic information via glutamate release pathway . In particular , astrocyte-mediated regulation of synaptic release could greatly increase paired-pulse facilitation ( PPF ) at otherwise depressing synapses ( and thus switch the synapse from depressing to facilitating ) ; conversely , it could reinforce paired-pulse depression ( PPD ) at otherwise facilitating synapses ( therefore switching the synapse from facilitating to depressing ) . These findings imply that astrocytes could dynamically control the transition between different “plasticity modes” . The present model also lends an explanation to several pieces of experimental data , as we detail below . In agreement with experimental results [26] , [27] , our model suggests that the type of presynaptic glutamate receptors targeted by astrocytic glutamate critically determines the type of modulation that takes place . The modulatory action of an astrocyte is lumped in our model into the so-called “effect” parameter α: lower values of α make the action of an astrocyte depressing with respect to the overall synaptic release but also increase paired-pulse facilitation . On the contrary higher α values make the effect of an astrocyte facilitating but at the same time paired-pulse depression is enhanced . Interestingly , some recent experiments on perforant path-granule cell synapses in the hippocampal dentate gyrus , show that facilitation of synaptic release mediated by astrocyte-derived glutamate correlates with a decrease of paired-pulse ratio [31] . Our model provides a natural explanation of these experimental results . Several lines of experimental evidence suggest that different types of glutamate receptors may be found at the same synaptic bouton [26] . The different types of receptors have different activation properties and hence could be recruited simultaneously or in a complex fashion [30] , [41] . Thus it is likely that α could take intermediate values between 0 and 1 . In one particular scenario , concurrence of astrocyte-mediated depression and facilitation could also lead these two effects to effectively cancel each other so that no apparent modulation of synaptic release is observed . Interestingly , in some recent studies , the Ca2+-dependent release of glutamate from astrocytes was reported not to affect synaptic function [69] , [70] , thus questioning the vast body of earlier experimental evidence pointing to the contrary . In our model we posit that an apparent lack of astrocytic effect on synaptic function could arise when the “effect” parameter α exactly matches the basal release probability of that presynaptic terminal , that is when α = U0* ( in which case equation 5 becomes U0 ( Γ ) = α , meaning that U0 does not depend on Γ anymore ) . This scenario would lead to concurrence of astrocyte-mediated depression and facilitation with no net observable effect on synaptic transmission . Whether de facto astrocytes decrease or increase synaptic release likely depends on the specific synapse under consideration and the functional implications that such different modulations could lead to [15] , [21] , [30] . In the former case for example , enhanced PPF could be not functionally relevant if release of neurotransmitter is strongly reduced by astrocyte glutamate signaling . In such situation the astrocyte would essentially shut down synaptic transmission , hindering the flow of information carried by presynaptic spikes [71] . On the other hand , for astrocyte-induced facilitation , an increase of released neurotransmitter could correspond to a similar increase of transmission probability [72] . However , the associated modulations of paired-pulse plasticity could also account for complex processing of specific – i . e . temporal vs . rate – features of input spike trains [2] , [51] , [57] that could not otherwise be transmitted by the single synapse , that is without the astrocyte . In a recent line of experiments on frog neuromuscular junction , it was observed that glial cells could govern the outcome of synaptic plasticity based on their ability to bring forth variegated Ca2+ dynamics [40] , [41] . In other words , different patterns of Ca2+ oscillations in perisynaptic glia were shown to activate different presynaptic receptors and thus to elicit different modulatory effects on neurotransmitter release [41] . This scenario would call for a future modification of our model to include a dependence on astrocytic Ca2+ dynamics of the effect parameter α . Nevertheless such observations are generally bolstered by our study . Our model predicts the existence of a threshold frequency for Ca2+ oscillations below which PPD ( PPF ) is predominant with respect to PPF ( PPD ) and above which the opposite occurs . This supports the idea that different spatiotemporal Ca2+ dynamics in astrocytes , possibly due to different cellular properties [73]–[76] , could provide specialized feedback to the synapse [40] . Moreover , our model suggests that different types of presynaptic glutamate receptors might not be necessary to trigger different modulations of synaptic transfer properties . The fact that the frequency of Ca2+ oscillations could bias synaptic paired-pulse plasticity subtends the notion that not only the nature of receptors , but also the dynamics of their recruitment by gliotransmitter could be a further critical factor in the regulation of synaptic plasticity [27] , [41] . This latter could eventually be dictated by the timing and the amount of released glutamate [27] , [44] as well as by the ultrastructure of astrocytic process with respect to synaptic terminals which defines the geometry of extracellular space [18] , [77] thus controlling the time course of glutamate therein [78] . Remarkably , the threshold frequency of Ca2+ oscillations that discriminates between PPD and PPF falls , in our analysis , within the range <2 . 5 min−1 of spontaneous Ca2+ oscillations displayed by astrocytes in basal conditions independently of neuronal activity [32] , [49] , [50] , [79] , [80] , hinting a possible role for these latter in the regulation of synaptic physiology . Spontaneous Ca2+ oscillations can indeed trigger astrocytic glutamate release [32] , [79]–[82] which could modulate ambient glutamate leading to tonic activation of presynaptic receptors [26] , [83] . In this fashion , spontaneous glutamate gliotransmission could constitute a mechanism of regulation of basal synaptic release . Notably , in a line of recent experiments , selective metabolic arrest of astrocytes was observed to depress Schaffer collateral synaptic transmission towards increasing PPF , consistently with a reduction of the basal synaptic release probability as predicted by our analysis [49] . The latter could be also relevant in the homosynaptic case of astrocytic glutamate exocytosis evoked by basal activity of the same presynaptic terminal that it feeds back to [21] , [27] , [84] . In such conditions , the ensuing influence of astrocytic glutamate on synaptic release correlates with the incoming synaptic stimulus also through Ca2+ dynamics in the astrocyte [85] , unraveling potentially new mechanisms of modulation of synaptic transmission and plasticity . Although we focused on regulation of astrocyte at single synapses , our analysis could also apply to synaptic ensembles [51] , [54] that could be “contacted” either by the same astrocytic process [30] , [32] , [80] or by different ones with locally synchronized Ca2+ dynamics [81] . In this case , modulation of the release probability by the astrocyte would support the existence of “functional synaptic islands” [86] , namely groups of synapses , intermittently established by different spatiotemporal Ca2+ dynamics , whose transmission mode and plasticity share common features . The implications that such dynamic astrocyte-synapse interaction might have with regard to information flow in neural circuitry remain to be investigated . Due to their capacity to modulate the characteristics of synaptic transmission , astrocytes could also alter the temporal order of correlated pre- and postsynaptic spiking that critically dictates spike-timing dependent plasticity ( STDP ) at the synapse [87] . Thus , astrocyte modulation of short-term plasticity could potentially contribute to ultimately shape persistent modifications of synaptic strength [49] , [88] , [89] underlying processing , memory formation and storage that provides the exquisite balance , subtlety and smoothness of operation for which nervous systems are held in awe [90] . Future combined physiological and computational studies will determine whether or not this is the case . | Synaptic plasticity is the capacity of a preexisting connection between two neurons to change in strength as a function of neuronal activity . Because it admittedly underlies learning and memory , the elucidation of its constituting mechanisms is of crucial importance in many aspects of normal and pathological brain function . Short-term presynaptic plasticity refers to changes occurring over short time scales ( milliseconds to seconds ) that are mediated by frequency-dependent modifications of the amount of neurotransmitter released by presynaptic stimulation . Recent experiments have reported that glial cells , especially hippocampal astrocytes , can modulate short-term plasticity , but the mechanism of such modulation is poorly understood . Here , we explore a plausible form of modulation of short-term plasticity by astrocytes using a biophysically realistic computational model . Our analysis indicates that astrocytes could simultaneously affect synaptic release in two ways . First , they either decrease or increase the overall synaptic release of neurotransmitter . Second , for stimuli that are delivered as pairs within short intervals , they systematically increase or decrease the synaptic response to the second one . Hence , our model suggests that astrocytes could transiently trigger switches between paired-pulse depression and facilitation . This property explains several challenging experimental observations and has a deep impact on our understanding of synaptic information transfer . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"circuit",
"models",
"neurophysiology",
"synapses",
"biophysic",
"al",
"simulations",
"computational",
"neuroscience",
"physiology",
"biology",
"anatomy",
"and",
"physiology",
"neurological",
"system",
"neuroscience",
"computational",
"biology"
] | 2011 | A Tale of Two Stories: Astrocyte Regulation of Synaptic Depression and Facilitation |
Recent genome wide association studies have identified a number of genes that contribute to the risk for coronary heart disease . One such gene , TCF21 , encodes a basic-helix-loop-helix transcription factor believed to serve a critical role in the development of epicardial progenitor cells that give rise to coronary artery smooth muscle cells ( SMC ) and cardiac fibroblasts . Using reporter gene and immunolocalization studies with mouse and human tissues we have found that vascular TCF21 expression in the adult is restricted primarily to adventitial cells associated with coronary arteries and also medial SMC in the proximal aorta of mouse . Genome wide RNA-Seq studies in human coronary artery SMC ( HCASMC ) with siRNA knockdown found a number of putative TCF21 downstream pathways identified by enrichment of terms related to CAD , including “vascular disease , ” “disorder of artery , ” and “occlusion of artery , ” as well as disease-related cellular functions including “cellular movement” and “cellular growth and proliferation . ” In vitro studies in HCASMC demonstrated that TCF21 expression promotes proliferation and migration and inhibits SMC lineage marker expression . Detailed in situ expression studies with reporter gene and lineage tracing revealed that vascular wall cells expressing Tcf21 before disease initiation migrate into vascular lesions of ApoE-/- and Ldlr-/- mice . While Tcf21 lineage traced cells are distributed throughout the early lesions , in mature lesions they contribute to the formation of a subcapsular layer of cells , and others become associated with the fibrous cap . The lineage traced fibrous cap cells activate expression of SMC markers and growth factor receptor genes . Taken together , these data suggest that TCF21 may have a role regulating the differentiation state of SMC precursor cells that migrate into vascular lesions and contribute to the fibrous cap and more broadly , in view of the association of this gene with human CAD , provide evidence that these processes may be a mechanism for CAD risk attributable to the vascular wall .
Atherosclerotic coronary heart disease ( CAD ) continues to be the dominant medical problem in the Western world and is growing in minority groups in this country and in developing populations [1–4] . Despite extensive investigation of epidemiological and cellular features of the disease , there are still fundamental questions related to the mechanism of etiology that remain to be answered . For instance , the resident SMC that constitute the majority of the native vessel wall undergo dramatic phenotypic changes in the disease setting , with the loss of SMC gene expression in the medial vessel layer and appearance of a new layer of SMC marker expressing cells in a “fibrous cap” that covers the lipid core of the plaque . While the risk of plaque rupture appears to be inversely correlated with the number of SMC-like cells in the fibrous cap , there is very little understanding of their in vivo origin , and the molecular pathways that regulate their expansion and terminal phenotype determination [5] Significant new insights into the fundamental cellular processes that drive CAD and other complex human diseases have recently been achieved through genome wide association ( GWA ) studies employing large cohorts of patients and healthy controls . These studies have provided the first incontrovertible identification of genes and disease pathways that affect risk for many complex human diseases ( www . genome . gov/gwastudies ) , including CAD . A recent GWA meta-analysis has identified 46 loci that associate with CAD at the genome-wide significance level of 5 x 10-8 , and another 104 independent variants that associate at a false discovery rate less than 0 . 05 [6] . Of these CAD associated loci , more than two-thirds act independently of traditional risk factors , and are likely to modulate risk of CAD through regulation of cellular processes in the blood vessel wall . Further study of this latter group of genes is expected to provide insights into novel atherosclerosis pathways . One CAD associated gene that has not been linked to known environmental or metabolic risk factors is the basic-helix-loop-helix transcription factor TCF21 ( capsulin , POD1 , epicardin ) [6–8] . This gene was initially cloned in four independent laboratories and shown to mark expression of mesodermal cells that give rise to kidney , lung , spleen , gonads , and a restricted group of cranial muscle cells [9–16] . Recent studies have shown that Tcf21 regulates fundamental cell fate decisions in the developing epicardium , serving as a determining factor for divergence between coronary vascular smooth muscle cell and cardiac fibroblast lineages [17 , 18] . In this setting , Tcf21 is downregulated in cells that are fated to become differentiated coronary SMC and has sustained expression in cells that become interstitial and adventitial fibroblasts [18] . However , its role in molecular and cellular processes related to CAD has not been investigated . To examine the possible role of TCF21 in coronary artery disease , we assessed its expression pattern in normal and diseased tissues , investigated its role in fundamental cellular processes , and employed next generation sequencing based methods in SMC to identify genes and pathways regulated by TCF21 at the transcriptional level . We have also identified the temporal and cell-specific expression of Tcf21 in reporter gene and lineage tracing mouse models to better understand how this gene might regulate vascular wall cellular processes in the setting of atherosclerotic disease .
To better understand the expression pattern of TCF21 in adult tissues , we studied expression in a number of adult human cells by quantitative real-time polymerase chain reaction ( qPCR ) ( S1A Fig ) . TCF21 was expressed in human coronary artery smooth muscle cells ( HCASMC ) , but not aortic smooth muscle cells ( HAoSMC ) or endothelial cells isolated from coronary ( HCAEC ) or aortic vessels ( HAoEC ) . Because of the reported similarity between pericytes , which have been developmentally linked to Tcf21 [10] , and mesenchymal stem cells ( MSC ) [19] , we investigated expression in MSC derived from bone marrow ( bmMSC ) , embryonic-stem cells ( eMSC ) and induced pluripotent stem cells ( iMSC ) . Results showed that all of these cells expressed TCF21 . Cardiac fibroblasts ( HCF ) and adventitial aortic fibroblasts ( HAoAF ) were also found to express TCF21 . Immunocytochemistry of cultured HCASMC revealed nuclear expression of TCF21 in cells that were also expressing the SMC marker ACTA2 , although a number of prominent cells expressing high levels of ACTA2 appeared to have low-level expression of TCF21 ( S1B Fig ) . In situ hybridization of mouse and human coronary artery specimens with species-specific cRNA probes revealed prominent labeling of adventitial cells with no staining of medial SMC or intimal endothelial cells ( S1C Fig ) . In preparation for studies in mouse genetic disease models , we investigated Tcf21 expression in normal adult mice with Xgal staining of tissues containing a lacZ reporter gene integrated into the murine Tcf21 locus , and compared this staining with that seen in C57BL/6 mice [12 , 20] . As with in situ hybridization studies in mouse and human tissues , results showed staining of cells in the adventitia of both the coronary arteries and aorta , with primary localization in those adventitial cells juxtaposed to the vascular media ( S1D Fig ) . Unique to the mouse , there was a patchy distribution of lacZ expressing cells in the proximal aortic wall ( S1E Fig ) . When β-galactosidase activity was pseudocolored green and merged with the Acta2 immunoreactivity ( red ) there appeared to be minimal overlap in either the coronary arteries or aorta . To provide functional insights into the role of TCF21 in adult SMC in the disease setting , we employed whole transcriptome RNA sequencing ( RNA-Seq ) studies on cells transfected with either non-silencing control siRNA ( siCTRL ) or silencing TCF21 siRNA ( siTCF21 ) . Quantitative RT-PCR of HCASMC transfected with siTCF21 compared to siCTRL showed a significant decrease in mRNA levels for TCF21 ( 0 . 99±0 . 02 control vs . 0 . 25±0 . 02 siTCF21 , P<0 . 0001 ) ( S2A Fig ) , and quantitation of western blots of protein extracts showed a similar reduction of TCF21 protein levels to 26% of baseline ( S2B Fig ) . We employed two commonly used algorithms , DESeq and edgeR , to identify genes that were differentially expressed ( FDR ≤ 0 . 05 ) between the HCASMC treated with siTCF21 or siCTRL reagents [21 , 22] . To eliminate outliers that might be identified by each analysis , we intersected the 466 genes obtained with DESeq and 430 genes found with edgeR to obtain 380 genes that were common to the two approaches [23] . The intersected genelist was mapped to the Ingenuity IPA Knowledge Base , generating algorithmically computed pathways or networks that were then investigated for over-representation in disease and biological function categories . “Connective Tissue Development and Function , Tissue Morphology , Cardiovascular Disease” was identified as the top network . “Cardiovascular Disease” was ranked third in the disease category , with 20 subcategories including “Vascular disease” ( Q = 3 . 26E-17 , right-tailed Fisher’s Exact Test , Benjamini-Hochberg corrected ) , “Disorder of artery” ( Q = 4 . 04E-12 ) , and “Occlusion of artery” ( Q = 2 . 03E-14 ) as the top three functional groups ( Tables 1 and S1 ) . “Cardiovascular System Development and Function” ( Q = 1 . 88E-16- 5 . 30E-04 ) was the most highly ranked in the “Physiological System Development and Function” category , with a number of subcategories that were related to embryonic development , including “Development of the cardiovascular system” ( Q = 1 . 88E-16 ) , “Angiogenesis” ( Q = 4 . 58E-14 ) , and “Development of blood vessel” ( Q = 4 . 58E-14 ) being the top three functional groups ( Tables 1 and S2 ) . Among the Molecular and Cellular Functions category of traits , the top three functional groups were “Cellular movement” ( Q = 4 . 23E-23 ) , followed by “Cellular growth and proliferation” ( Q = 4 . 23E-23 ) , and “Cell death and survival” ( Q = 1 . 06E-18 ) ( Tables 1 and S3 ) . Taking the first subcategory in the “Cardiovascular Disease” category , “Vascular disease , ” we used well-curated molecular interactions in the IPA Knowledge Base to build a gene network . The initial group included 73 genes , and 18 additional genes were added with the “build” function in IPA , this group functionally related with the “connect” function , and with elimination of non-connected nodes resulted in a final TCF21 Vascular Disease Network containing 83 genes ( Fig 1 ) . Also , content was added to the network with annotation from Gene Ontology classifications ( S3 Fig ) . One of the strongest signals in the network was the upregulation of matrix components with TCF21 knockdown , including a large number of collagens , SPARC , LOX , vitronectin ( VTN ) , and components of the matrix that are known to affect vascular cell functions , including thrombospondin 1 ( THBS1 ) and fibrillin 1 ( FBN1 ) , as well as vascular cell surface proteins that interact with the matrix including alpha V integrin ( ITGAV ) . Although the matrix degrading enzymes MMP7 , MMP10 and MMP25 are upregulated with TCF21 knockdown , the downregulation of the much more highly expressed MMP1 and MMP3 combined with the upregulation of MMP inhibitor TIMP3 are consistent with an anti-matrix overall profile of a mobile cell . Also upregulated are SMC markers such as smooth muscle actin ( ACTA2 ) and myosin light chain kinase ( MYLK ) . Among the genes downregulated with TCF21 knockdown are those involved in vascular development , including neuropilin 1 ( NRP1 ) , VEGF receptors II ( KDR ) , angiopoietin 1 ( ANGPT1 ) , TGFβ receptor 2 ( TGFBR2 ) , and semaphorin 3D ( SEMA3D ) , and cytokines and chemokines such as IL1A , IL1B , CXCL2 and CXCL3 . Taken together these data suggest that TCF21 positive cells express low levels of many matrix proteins while expressing high levels of matrix degrading enzymes , consistent with a pro-migratory phenotype , express low levels of SMC lineage markers , and are highly pro-angiogenic and pro-inflammatory , a phenotype quite different from vascular wall SMC . To investigate a functional role for TCF21 in fundamental cell fate decisions in vascular SMC , we employed lentiviral shRNA as well as siRNA mediated knockdown and lentiviral over-expression studies . Control lentiviral vectors ( pWPI ) and lentiviral over-expression vectors ( pWPI-TCF21 ) , and control ( pLVTHM ) and lentiviral shRNA mediated knockdown vectors ( pLVTHM-shTCF21 ) were used to transduce primary cultured HCASMC . pWPI-TCF21 increased TCF21 mRNA levels ( 1 . 0±0 . 04 pWPI vs . 32 . 5±0 . 02 pWPI-TCF21 , P<0 . 0001 ) , and pLVTHM-shTCF21 decreased expression ( 1 . 0±0 . 06 pLVTHM vs . 0 . 34±0 . 04 pLVTHM-shTCF21 2 , P<0 . 001 ) ( S4A Fig ) . Western blot of protein extracts from HCASMC that were transduced with over-expression and knockdown lentiviruses showed a 4 . 5-fold increase , and reduction of TCF21 protein levels to 8% ( shTCF21 1 , shTCF21 2 ) of baseline respectively ( S4B Fig ) . To assess the effect of TCF21 on rate of cell division in vitro , proliferation rates of pools of stably transduced and non-transduced cells were measured over time , where increased or decreased proliferation rates in transduced cells would lead to a change in percentages of this GFP positive cell population as compared to the non-transduced , GFP negative population . Flow cytometry analysis of stable overexpression of TCF21 in HCASMC showed an increase in TCF21 overexpressing cells from 48 to 82 percent of the culture within 28 days ( S5A Fig ) . In the same time period the percentage of cells with stable knockdown of TCF21 decreased from 53 to 16 for construct 1 ( shTCF21 1 vs . empty control virus pLVTHM P<0 . 0001 ) and from 59 to 2 percent for construct 2 ( shTCF21 2 vs . empty control virus pLVTHM P<0 . 0001 ) ( S5B Fig ) . The percentage of GFP positive cells did not change when transduced with the empty vector control construct . To provide additional support for these findings we conducted EdU ( 5-ethynyl-2’-deoxyuridine ) labeling proliferation assays . HCASMC transduced with the control , overexpression or knockdown lentiviruses were treated with EdU and imaged for nuclear fluorescence . These data showed an increase in the percentage of TCF21 overexpressing cells compared to DAPI stained cells ( 33 . 1% ± 2 . 3 control vs . 51% ± 4 . 1 overexpressing cells , P<0 . 001 ) , and significant decreased percentage of dividing cells transduced with knockdown lentiviruses ( 43 . 4% ± 4 . 4 control vs . 30 . 8% ± 2 . 4 knockdown cells , P<0 . 01 ) ( Fig 2A and 2B ) . These data suggest that TCF21 positively affects SMC proliferation . We also evaluated the effect of TCF21 on apoptosis and migration in cultured HCASMC . Apoptosis was evaluated in HCASMC transfected with siControl or siTCF21 RNAs and serum starved for 48 hrs . By caspase assay there was a significant decrease in apoptosis of cells transfected with the siTCF21 reagent ( 41 , 824±1872 vs . 18 , 837±1302 for siCTRL compared to siTCF21 , P<0 . 0001 ) ( Fig 2C ) . For migration , HCASMC transduced with TCF21 overexpression lentivirus were evaluated with a gap closure assay . Here , TCF21 overexpressing cells covered a significantly larger surface area after 12 hours of incubation ( 27 . 8 ± 2 . 7 vs . 15 . 8 ± 1 . 16 control cells , P<0 . 001 ) ( Fig 2D ) . With an average doubling time of 71 hours for TCF21 overexpressing HCASMC this effect could not be due to the pro-proliferative effect of TCF21 . Given that TCF21 is downregulated in epicardial cells as they differentiate into coronary SMC , and a number of SMC lineage marker genes were upregulated in the siTCF21 RNA-Seq study , including ACTA2 and MYLK , we specifically investigated the expression of SMC marker genes in additional siRNA knockdown experiments . These studies showed that ACTA2 , TAGLN and MYH11 transcript levels were significantly upregulated after 70% TCF21 knockdown ( 0 . 96 ± 0 . 023 vs . 1 . 38 ± 0 . 16 , P<0 . 05 , 1 . 0 ± 0 . 03 vs . 2 . 59 ± 0 . 27 , P<0 . 01 , and 1 . 05 ± 0 . 09 vs . 2 . 67 ± 0 . 51 , P<0 . 05 , respectively ) ( Fig 2E ) , suggesting their direct suppression by TCF21 at the transcriptional level . To investigate whether TCF21 directly binds and regulates SMC marker genes , a series of studies were conducted with the ACTA2 gene as a representative SMC marker . It is well known that the regulatory regions of the ACTA2 gene are localized in the upstream promoter and first intronic regions of the gene [24] , and this information in conjunction with available ENCODE histone modification data , ENCODE DNase hypersensitivity data from human aortic smooth muscle cells , and localization of E-box binding motif sequences provided the approach for choosing a reporter construct for these studies ( Fig 2F ) . Chromatin immunoprecipitation was employed to verify the in vivo binding of TCF21 to a region of the first intron enhancer identified through this analysis ( Fig 2G ) . The luciferase reporter gene construct employed for these studies contained the human ACTA2 promoter and first intron . Obligate heterodimer partners of TCF21 , TCF3 and TCF12 , were included in these experiments to facilitate transcriptional activation [25] . Also , based on the high degree of conservation of sequence in the bHLH domain between TCF21 and Twist , and the similar inhibitory developmental actions of this related transcription factor , an expression construct for Twist was included to investigate possible interactions with TCF21 in the context of ACTA2 reporter gene expression [26] . When this reporter was transfected into rat aortic smooth muscle cells ( RASMC ) in conjunction with a TCF21 expression construct , it was found to decrease expression ( 1 . 0 ± 0 . 01 vs . 0 . 1 ± 0 . 009 , P<0 . 05 ) , and even greater reductions were observed when the TCF21 expression vector was combined with another vector expressing one of the class II bHLH factors TCF3 ( protein isoforms E12 or E47 ) ( Fig 2H ) . TCF21 was also shown to inhibit expression of a minimal luciferase reporter containing 3 E-box sequences based on the known TCF12 motif ( CAGCTG ) in conjunction with a minimal promoter . There was significant repression with TCF21 expression alone ( 1 . 0 ± 0 . 29 vs . 0 . 21 ± 0 . 02 , P<0 . 05 ) , and additional reductions were noted when bHLH class II factors were included ( Fig 2I ) . For both constructs , the developmental factor Twist produced a modest increase in TCF21 transrepression . To investigate disease-related expression of Tcf21 in the vascular wall , we performed time-course studies in Tcf21lacZ/+ , ApoE-/- hyperlipidemic atherosclerotic mice , and compared Xgal staining in these tissues with that observed in negative control ApoE-/- animals . After four weeks of high fat diet ( HFD ) there were no β-galactosidase expressing cells in the forming lesions , although there was apparent clustering of β-galactosidase positive cells on the luminal side of the media immediately below the areas where lesions had initiated growth ( Fig 3 ) . By eight weeks of HFD , lesions in most animals showed abundant β-galactosidase positive cells accumulating toward the luminal side of the atheromatous lesion , with streaks extending from media through the lesion toward the fibrous cap . β-galactosidase positive cells observed in the lesion appeared to be aligned orthogonal to the axis of orientation for cells in the media , suggestive of migratory behavior . By twelve weeks of HFD feeding there was extensive staining in lesions of most animals , with the β-galactosidase positive cells in the vicinity of the forming fibrous cap , or in a cap-like structure that was situated medial to and extended below the actual fibrous cap ( Figs 3 and 4 ) . This time interval from 4 to 12 weeks was associated with an apparent decrease in the number of medial cells expressing SMC markers and a disruption of the lamellar medial structure . Tcf21 expressing cells in the media at 8 to 12 weeks were disorganized and had lost their characteristic lamellar structure . By 20 weeks , almost all of the Tcf21 expressing cells in the lesions were associated with the fibrous cap or a subcapsular structure ( Fig 3 ) . To determine whether results in the ApoE model were consistent with TCF21 expression in the diseased human coronary vessel wall , immunostaining of human coronary artery tissues was performed with antibodies for TCF21 and ACTA2 . In those vessels with only modest disease there was adventitial TCF21 staining with some staining of the minimal neointima , but no TCF21 staining in the media , which was positive for ACTA2 ( S6A–S6C Fig ) . In vessels with significant disease there was robust TCF21 staining in the adventitia and neointima and some patchy staining in the fibrous cap , but very little staining in the media ( S6D and S6E Fig ) . ACTA2 staining was prominent in the fibrous cap and considerably weaker staining in the media . These studies confirm what was observed in the mouse diseased vasculature . To assign TCF21 expressing cells to a known cellular lineage we evaluated combined lacZ expression with immunohistochemistry staining for well-known lineage markers for different vascular cell types . Specifically , we worked to identify co-staining for murine adventitial and vascular lesion lineage markers , focusing on lacZ reporter gene expression in the ApoE-/- mice . Xgal and immunohistochemistry staining was performed of adventitial and lesion areas with antibodies specific for endothelial cells ( VWF ) , SMC ( ACTA2 ) , neuronal cells ( TUBB3 ) , adipocytes ( PLIN1 ) , fibroblasts ( THY1 ) , stem cells ( nestin ) , macrophages ( LAMP2 ) , hematopoietic cells ( PTPRC ) , and dendritic cells ( Itgax ) ( S7 Fig ) . These studies failed to identify co-localization with known cellular lineage markers , and when combined with published negative data for pericyte markers argue that Tcf21 is not expressed by a common terminally differentiated vascular cell type [17] . Further reporter gene and immunohistochemistry studies focused on the fibrous cap in lesions of mice fed HFD for 4 to 20 weeks . Studies with SMC marker Tagln again revealed that the majority of Tcf21 expressing cells in the adventitia , media and plaque of diseased vessels did not express SMC markers , which were easily identified in association with medial and fibrous cap SMC ( Fig 4 ) . Although individual Tcf21 expressing cells and those in subcapsular structures did not express SMC markers , there were Tcf21 expressing cells identified immediately below or within the fibrous cap that appeared to also express Tagln . In addition to SMC lineage markers , we investigated expression of growth factor receptor Tgfbr2 , which is involved in SMC development and structural vascular wall disease [27 , 28] and was identified as one of the differentially expressed genes in the TCF21 knockdown RNA-Seq studies , as well as Pdgfrb which has an important role in coronary SMC development and atherosclerotic disease [5 , 29] . Fibrous cap cells were shown to express growth factor receptors Tgfbr2 and Pdgfrb , and there appeared to be co-staining for Tcf21 expression and growth factor expression by cells in the vicinity of the fibrous cap . Whether the observed co-staining identifies a unique transitional cell that expresses Tcf21 and SMC marker genes , or is due simply to sustained enzymatic activity of the β-galactosidase protein after Tcf21 expression is downregulated was not clear from these studies . However , these data suggest that Tcf21 expressing cells can adopt an SMC phenotype and contribute to the fibrous cap . In the next series of experiments we sought to more clearly identify the possible origins and cellular fate of Tcf21 expressing cells found in the vascular disease lesions . For these studies we employed a lineage tracing approach with an inducible Cre recombinase that provided permanent fluorophore marking of cells that expressed Tcf21 at the time of activation of recombinase expression . We used a mouse line expressing a Cre recombinase protein fused to two mutant estrogen-receptor ligand-binding domains ( MerCreMer ) [30] , under the control of the endogenous Tcf21 locus ( Tcf21iCre/+ ) . This line has been well characterized and used extensively to investigate the cellular fate of Tcf21 expressing cells in the developing embryo [17 , 31] . For reporter alleles , we employed both a tandem dimer tomato ( tdT ) fluorescent reporter line ( B6 . Cg-Gt ( ROSA ) 26Sortm14 ( CAGtdTomato ) Hze/J ) as well as a lacZ reporter line ( B6 . 129S4-Gt ( ROSA ) 26Sortm1Sor/J ) that was tracked with histochemical β-galactosidase activity . ApoE-/- mice carrying the Cre and reporter alleles were administered tamoxifen before HFD and sacrificed after 16 weeks of the diet , or administered tamoxifen at 6–8 weeks and harvested after 8 , 12 or 16 weeks of HFD before sacrifice . Lineage tracing studies in ApoE-/- mice were primarily conducted in animals on HFD for 12 weeks , with tamoxifen being administered from 6 to 8 weeks of the diet . At baseline , Tcf21 expression using this lineage tracing system was consistent with that seen for the lacZ knockin reporter mouse line , being restricted to the adventitial cells associated with coronary arteries , cardiac fibroblasts , and some aortic root cells ( Fig 5A ) . Aortic root tissues from animals treated with HFD revealed tdT-expressing cells in the lesions . For early lesions without a defined fibrous cap , Tcf21 lineage traced cells were distributed through the lesion and did not express the SMC marker Tagln ( Fig 5B ) . In those lesions with a mature fibrous cap , Tcf21 lineage traced cells were noted to contribute to this cap and many of these cells shown to express Tagln ( Fig 5C and 5D ) . These data are consistent with the results described above for Tcf21 reporter gene and SMC marker immunohistochemistry studies , and suggest that Tcf21 expressing cells may give rise to SMC in the fibrous cap . By activating recombinase activity before HFD feeding and lesion development it was possible to investigate the cellular origin of the Tcf21 expressing cells identified in the vascular lesions of the ApoE-/- model . For mice that were administered tamoxifen before HFD , and then maintained on the diet for 12 weeks , lineage traced cells were commonly identified in lesions ( Fig 5E ) . These data are thus consistent with their derivation from Tcf21 expressing cells in the vascular wall adventitia or media , and suggest that Tcf21 lineage traced cells were able to migrate into the forming lesions and contribute to lesion formation . Finally , to verify the results obtained with the fluorescent tdT reporter , we conducted identical experiments with the Cre-activated lacZ reporter mouse . These studies also found that Tcf21 lineage traced cells in the neointima did not express the SMC marker Tagln , whereas lineage traced cells contributing to the fibrous cap cells were found to express this SMC marker ( S8 Fig ) . To generalize the findings obtained in the ApoE-/- model , we conducted lineage tracing studies in Ldlr-/- mice . Findings from these experiments also showed that activation of the reporter gene at 3 to 5 weeks of age , prior to HFD feeding for 16 to 20 weeks , marked cells that migrated into the atherosclerotic lesion ( Fig 6A ) . As in the ApoE-/- model , some of the Tcf21 traced cells activated expression of SMC markers Tagln and Acta2 and localized to the fibrous cap ( Fig 6B ) . In addition , co-labeling with an antibody specific for the matrix protein periostin showed some general co-localization of Tcf21 traced cells and periostin in the adventitia while there was extensive colocalization in the fibrous cap and co-staining in a subcap region located abluminal to the fibrous cap ( Fig 6C ) . Also , Tcf21 traced cells in the adventitia and fibrous cap showed extensive colocalization for Pdgfra expression ( Fig 6D ) . To investigate whether Tcf21 plays a role in regulating vascular cell proliferation in the disease setting in vivo , we used two different approaches , one employing early stage disease in the Ldlr-/- model , and the other a later stage ApoE-/- murine model . In experiments with Ldlr-/- mice we investigated whether there was a time-dependent increase in the relative number of Tcf21 expressing cells that were undergoing cell division . In these studies , Tcf21iCre/+ , ROSAtdT/+ , Ldlr-/- mice were treated with tamoxifen at 1 week before isolation , and EdU was injected 24 hours prior to tissue harvest at relatively early disease time points for this model , at 0 , 9 , 12 , and 15 weeks of feeding with HFD . EdU positive cells were found in the adventitia , media and intima of the aortic wall and quantified in sections of the aorta through the region of the sinus . The percentage of proliferating Tcf21 expressing cells was identified by both tdT and EdU fluorescence ( Fig 7A ) . Comparison of the number of Tcf21+EdU+ cells versus all Tcf21+ cells at different time points showed a significantly greater percentage of proliferating cells at 12 weeks of diet compared to time zero ( 1 . 94±0 . 70 vs . 6 . 42±1 . 10 , P<0 . 05 ) ( Fig 7B ) . There was also an increase in the mean percentage of EdU+ Tcf21+ cells to EdU+ cells between the 9 and 12 week time points , although this difference did not reach statistical significance ( P = 0 . 053 ) . Studies also investigated whether there was a difference in the relative rate of cell division in lesions of Tcf21lacZ/+ , ApoE-/- mice compared to ApoE-/- animals , after 20 weeks of HFD . All mice were injected with EdU 24 hours prior to tissue harvest and the number of EdU fluorescent labeled cells was compared to the total number of cells as identified by DAPI fluorescence for each genotype . This analysis was restricted to the plaque area where the majority of EdU+ cells were located in these more mature lesions ( Fig 7C and 7D ) . The Tcf21lacZ/+ , ApoE-/- mice showed a significantly lower number of EdU+ cells compared to ApoE-/- animals ( 2 . 25±0 . 18 vs . 3 . 27±0 . 25 , P<0 . 05 ) ( Fig 7E ) . Taken together , these data suggest that Tcf21 expressing cells show a time-dependent increase in proliferation with disease progression , possibly greater than other cells not expressing Tcf21 , and that this increase may depend on Tcf21 expression . To begin to investigate the possible association of TCF21 expression and function in the fibrous cap of human atherosclerotic lesions , we have investigated the TCF21 expression pattern in fibrous plaque of carotid atherosclerotic lesions collected from patients at time of endarterectomy surgery . Micro-dissection laser capture was used to obtain tissue from the fibrous cap of 10 vessels that showed stable plaque morphology and 10 vessels that had undergone plaque rupture ( Fig 8 ) . RNA was isolated from these tissues and employed to assess TCF21 mRNA levels by qPCR . This analysis showed significantly decreased levels of TCF21 expression in the fibrous cap tissue harvested from the ruptured plaque ( 5 . 0±0 . 3 vs 6 . 9±0 . 3 , ruptured vs stable plaque , P<0 . 0001 ) .
SMC are critical contributors to atherosclerosis pathophysiology , and relevant to the findings reported here is the dramatic alteration in gene expression and basic cellular functions that occur with phenotypic modulation of this cell type in the disease setting [5 , 32 , 33] . In the classic response to injury hypothesis , fully differentiated medial SMC express contractile marker genes and have limited proliferative capacity , while under disease conditions they downregulate contractile gene expression , proliferate , and migrate into the neointimal space . Data presented here suggest that upregulation of TCF21 may be directly responsible for at least some aspects of the phenotypic modulation that is believed to characterize the medial SMC response to vascular injury . We have shown that TCF21 binds the ACTA2 locus , transrepresses ACTA2 reporter genes in cultured SMC , and that knockdown of TCF21 is associated with increased ACTA2 expression . Although we have not studied other SMC markers in such detail , both TAGLN and MYH11 are upregulated in siTCF21 treated cells . While additional studies are required to characterize a possible dedifferentiation function for TCF21 , also consistent with such a hypothesis are in vitro studies in cultured HCASMC presented here which show that TCF21 promotes proliferation and migration . A differentiation inhibiting function has previously been proposed for TCF21 in skeletal muscle cells [34] . Tcf21 is expressed in epicardial progenitor cells that give rise to coronary SMC , and its expression in the adult might be characteristic of a precursor cell that is resident in the vascular wall and is capable of giving rise to SMC in the setting of vascular injury . A number of previous studies have identified such precursor cells in distinct compartments of the vessel wall , all with some capacity to give rise to differentiated SMC . It is generally accepted that SMC accumulation in the fibrous cap is monoclonal or oligoclonal , and this is consistent with the existence of a resident arterial subpopulation of SMC precursor cells that contribute to healing in the response to disease stresses [35] . In bovine and canine vessels , subsets of medial SMC have prominent proliferation with reduced SMC marker expression , and this population has been estimated to be 6–15% of cells in the murine aorta [36–38] . These observations are consistent with data presented here showing that Tcf21 expressing cells with low SMC marker expression are distributed in the proximal aorta in mouse . Cells with SMC precursor function have also been identified in the adventitial layer [35] . In mouse , Sca1+ cells with SMC differentiation potential have been identified in the aortic adventitia [39] , and in humans , a population of cells with SMC potential has been identified in the space between the media and adventitia of large and medium sized vessels [40] . While there are reported differences in the features and markers of these and other progenitor-like cells in the vascular wall , here we demonstrate an overlap of expression pattern and other features in Tcf21 expressing cells . It seems likely that TCF21 may be expressed in one or more of these cell populations and may contribute to an undifferentiated SMC precursor phenotype . The studies described herein provide important new information regarding the biology of the fibrous cap . While the risk of plaque rupture appears to be inversely correlated with the number of SMC in the fibrous cap , there is limited understanding of their in vivo origin , and the molecular pathways that regulate their expansion and terminal phenotype determination [5 , 41] . With these studies we have shown Tcf21 to be a marker for precursor cells that give rise to at least a portion of fibrous cap cells , and genomic studies in conjunction with in situ protein marker studies identified a number of signaling molecules that relate to the phenotype of this important cell type . While Pdgfra and Pdgfrb have been previously identified on fibrous cap cells [42 , 43] , we show here that they are not expressed by cells that are migrating toward the cap and expressing Tcf21 but initiate expression in conjunction with upregulation of SMC marker genes when they are juxtaposed to the endothelium . Tgfbr2 has not previously been identified on fibrous cap cells , and its biology in this context not studied , but it has recently been shown to support the differentiated SMC phenotype and homeostasis of this cellular lineage [27 , 28] . TGFBR2 mRNA levels were found to be regulated by TCF21 in the siRNA knockdown RNA-Seq studies . Another protein expressed by Tcf21 lineage-traced cells in the fibrous cap , periostin , is known to promote SMC differentiation and migration [44–46] , and interestingly has recently been associated with the atherosclerotic phenotype in young humans [47] . Additional follow-up of genomic studies are expected to provide insights into other genes and pathways that are present in TCF21 marked fibrous cap precursor cells . Lineage tracing provides a unique and informative approach to address questions related to cellular origin , movement and differentiation of lesion cells in the vascular wall , however a number of questions remain unanswered . Given the capacity of Tcf21 to promote autonomous cellular proliferation and migration , and the appearance of lineage traced cells in lesions of animals where recombination preceded HFD feeding , it seems likely that lineage traced cells can migrate from the adventitia or media to contribute to lesion formation . Although there is some evidence that bone marrow—derived SMC-like cells may contribute to neointima formation , such cells appear to represent a very small fraction of the total neointimal cells [48 , 49] . Given the variability in disease progression in different ApoE-/- and Ldlr-/- animals , and the possibility that Tcf21 is upregulated in adventitial or medial cells throughout disease progression , lineage tracing does not allow assessment of how much of the fibrous cap derives from Tcf21 expressing cells . Results from the Tcf21lacZ/+ , ApoE-/- studies suggest that Tcf21 expressing cells constitute a significant portion of the lesion SMC but this cannot be readily evaluated with current methods . Also , evidence that some fibrous cap SMC derive from lineage traced cells does not eliminate the possibility that a portion of the fibrous cap cells continue to express Tcf21 and maintain a fibroblast or myofibroblast phenotype , and that the fibrous cap is actually a combination of such cells along with those that have downregulated Tcf21 and adopted a true SMC phenotype . This possibility is consistent with sustained Tcf21 expression in some of the fibrous cap cells at the 20-week disease timepoint ( Fig 3 ) . While the work reported here has provided important initial details of how this gene is expressed during atherosclerotic vascular disease and the types of processes that it might regulate to affect disease pathophysiology , the causal mechanisms of disease risk remain unclear . The direction of effect has not been fully established by current available human genetic and genomic data , with both directions being suggested by different approaches and possible underlying gene regulatory mechanisms characterized for each [8 , 50 , 51] . Based on the animal model findings reported here , greater TCF21 expression could inhibit disease risk by promoting the expansion of SMC lineage and contributing to plaque stabilization by supporting fibrous cap development . Alternatively , greater TCF21 expression in SMC precursor cells at the fibrous cap might inhibit their differentiation to the SMC lineage , thus destabilizing the plaque . Similar alternative hypotheses could be generated for the risk allele producing decreased TCF21 expression . Nonetheless , further in vivo gain- and loss-of-function studies are required to determine how changes in Tcf21 gene expression affects the size and architecture of the fibrous cap and how such changes correlate to specific human disease traits linked to plaque vulnerability and rupture . These questions could be addressed using the various plaque rupture models established in mice [52 , 53] . Finally , it will be important to investigate the phenotype of the TCF21 expressing cells at baseline , and how the phenotype changes as these cells migrate into the lesion . Such studies could be performed with captured cells using modern genomic methods , and would provide extremely valuable insights into the cellular phenotype as these cells respond to the vascular signals of the disease process .
All procedures described in this study were approved by the Institutional Animal Care and Use Committees of Stanford University and University of Hawaii and conformed to NIH guidelines for care and use of laboratory animals . Human primary Coronary Artery Smooth Muscle Cells ( HCASMC , #CC-2583 ) were obtained from Lonza ( Allendale , NJ , USA ) at passage 4 and cultured in Smooth Muscle Growth Medium-2 including hEGF , insulin , hFGF-B and FBS , but without antibiotics ( Lonza , #CC-3182 ) . Human embryonic kidney cells ( HEK , ATCC #CRL-11268 ) and rat aortic smooth muscle cells ( A7R5 , ATCC #CRL-1444 ) were cultured in DMEM High Glucose ( Life Technologies #311995–065 ) with 10% FBS ( Life Technologies #26140–79 ) . Human cardiac fibroblast ( HCF ) , bone-marrow derived mesenchymal stem cells ( bmMSC ) , embryonic stem cell derived mesenchymal stem cells ( eMSC ) , induced pluripotent stem cell-derived mesenchymal stem cells ( iMSC ) , human aortic adventitial fibroblasts ( HAoAF ) , human aortic endothelial cells ( HAoEC ) , human coronary artery endothelial cells ( HCAEC ) , and human aortic smooth muscle cells ( HAoSMC ) were obtained from commercial suppliers , maintained in the recommended media , and evaluated with at least 3 technical replicates . For RNA-Seq studies and individual gene quantitative PCR , HCASMC were transfected with 300 nM TCF21 Trilencer-27 Human siRNA ( OriGene #SR304753C ) or Trilencer-27 Universal Scrambled Negative Control siRNA ( OriGene #SR30004 ) at 80% confluence using the Amaxa Basic Nucleofector Kit for Primary Mammalian Smooth Muscle Cells ( Lonza #VPI-1004 ) at a density of 1 x 106 cells per 100 μL sample using Nucleofector Program U-025 . Cells were changed to medium with supplements at 18 hours post-transfection and cultured for an additional 48 hours . For TCF21 overexpression , a cDNA clone ( OriGene #SC125048 ) representing the protein coding sequence of human TCF21 was cloned into pWPI vector ( Addgene #12254 ) which also contains the GFP reporter gene . For shRNA knockdown the following sequences were cloned into pLVTHM ( Addgene #12247 ) : For qPCR RNA was isolated using QIAzol ( Qiagen #79306 ) and the Trizol RNA isolation protocol . DNA was removed with the DNA-free kit from Ambion ( Life Technologies #AM1906 ) followed by first strand cDNA synthesis and subsequent RNA digestion according to the SuperScript III protocol ( Life Technologies #18080–051 ) . Relative gene expression was measured using the ViiA 7 Real-Time PCR System from Applied Biosystems ( Life Technologies ) . cDNA was amplified as follows: 95°C 15 min , 40x ( 95°C 1s , 60° 20s ) . Locus enrichment in chromatin immunoprecipitation samples was measured using the ABgene SYBR mastermix ( AB-1166 ) and the primers ACTA2chipfor 5’-AGGGAGATGCAAACCAGATATCC-3’ , ACTA2chiprev 5’-GCAGGTGACCTGCTGAATTTTTC-3’ , PROCRchipfor 5’-GACTCCTGCTTACCTCCTCATA-3’ and PROCRchiprev 5’-GGGTGAAGAAGGTACAAAAGAA-3’ . Genomic DNA cycling conditions were 95°C 15min , 40x ( 95°C 15s , 60°C 1 min ) . Absolute quantification of expression levels was performed using a standard dilution series and normalization to 18S rRNA . Relative expression levels were then calculated relative to TCF21 expression levels in HEK cells ( S1A Fig ) or in SiCTRL treated HCASMC ( Fig 2E ) by division . Total RNA from either siTCF21 or siCTRL treated samples was depleted for ribosomal RNA with the Ribo-Zero magnetic kit from Epicentre ( Illumina #MRZH116 ) , libraries generated with the Epicentre ScriptSeq v2 RNA-Seq library preparation kit ( Illumina #SSV21106 ) and thereafter sequenced as 100bp paired-end reads on an Illumina HiSeq 2000 instrument . The resulting data has been deposited at GEO under accession number GSE44461 . Reads resulting from RNA-Sequencing of siCTRL and siTCF21 treated HCASMC were mapped using software tools TopHat+Bowtie2 . Differential expression level between samples was analyzed using the software tools DESeq and edgeR at an FDR ≤ 0 . 05 , with intersection of the 466 and 430 respective identified genes providing a group of 380 common genes . This genelist was used to interrogate the Ingenuity Knowledge Base , identifying over-representation of Cardiovascular Disease annotation terms , including the top functional category “Vascular disease . ” Gene members of this group were expanded by adding 18 genes with the “build” function in IPA to create the TCF21 Vascular Disease Network . Visualization of the network was performed using Cytoscape open source software . Node color was mapped to log fold change ( green-yellow-red palette ) , node size to absolute expression value in wild type cells , and font size to enrichment Q-value . GO categories were assigned to genes using the Bingo application for Cytoscape and coloring the GO categories was performed using GOlorize application for Cytoscape . A Tcf21 functional knockout and reporter line , Tcf21lacZ/+ , was constructed by substituting the lacZ gene into the first exon of the murine gene employing homologous recombination in embryonic stem cells , as described previously [12] . Fidelity of expression of the lacZ reporter has been described [12] , and methods for using Xgal staining to track Tcf21 expression were identical to those published previously from this laboratory [54] . For studies of reporter gene expression in the ApoE-/- model , animals were weaned onto Western high fat diet ( HFD , 21% anhydrous milk fat , 19% casein and 0 . 15% cholesterol , Dyets no . 101511 ) . The animals were euthanized and perfused with PBS followed by 0 . 4% PFA , the aortic roots excised and subjected to Xgal staining followed by post-fixation for 30 minutes in 4% PFA . After incubation in an ascending sucrose solution series and embedding in OCT , tissue was processed into 7 μm thick frozen sections for immunohistochemistry . Lineage tracing studies employed a recently described inducible Tcf21-Cre line ( Tcf21iCre/+ ) [31] , and tomato ( B6 . Cg-Gt ( ROSA ) 26Sortm14 ( CAGtdTomato ) Hze/J ) Jax #007914 and lacZ ( B6 . 129S4-Gt ( ROSA ) 26Sortm1Sor/J ) Jax # 003474 reporter lines . The Tcf21iCre/+ line was obtained on a mixed C57Bl/6 x 129SV mixed background and was bred back onto the C57Bl/6 background for 6 generations before being mated with the ApoE-/- line . The Tcf21iCre/+ allele was bred directly onto the Ldlr-/- line in C57Bl/6 background . All Ldlr-/- mice were treated with a modified Western high fat diet ( HFD , 15 . 8% milk fat and 1 . 25% cholesterol ( diet 94059; Harlan Teklad ) for the time periods indicated . Tcf21iCre/+ , ROSAtdT/+ , ApoE-/-; Tcf21iCre/+ , ROSAtdT/+ , Ldlr-/-; and Tcf21iCre/+ , ROSAlacZ/+ , ApoE-/- and Ldlr-/- and ApoE-/- control animals were treated as described with HFD , tissues harvested and sectioned as described above , and analyzed for expression of the reporter genes as well as for other cellular markers as described below . For ApoE-/- animals , Cre recombination was activated with a series of ten 1 mg tamoxifen ( Sigma Aldrich ) intraperitoneal injections for a total of 10 mg of tamoxifen . Animals were induced at 0 , 6 , 10 weeks and sacrificed after 20 , 8 , and 12 weeks respectively of the HFD as described above . For Ldlr-/- mice , Cre induction was accomplished by feeding of tamoxifen chow at 3–5 weeks of age ( Harlan Laboratories , Inc ) . For immunocytochemistry , HCASMC were seeded into culture slides ( BD Biosciences #35641 ) in supplemented medium for 24 hours . For immunohistochemistry of human tissues , paraformaldehyde-fixed human coronary arteries were frozen , embedded and sectioned at a thickness of 7 μm and tissue slides were postfixed with 4%PFA in PBS for 10 minutes . Staining for TCF21 expression was conducted on frozen tissue sections with a polyclonal anti-TCF21 antibody ( Sigma-Aldrich #HPA013189 ) at a dilution of 1:100 , with overnight primary incubation . Mouse tissues were collected and processed as described above . After postfixation , nonspecific binding sites were blocked for 30 minutes with Rodent Block M ( Biocare Medical #RBM961 ) . Primary antibody incubation was performed overnight at 4°C using the following antibodies and dilutions in Da Vinci Green antibody diluent ( Biocare Medical #PD900 ) , anti-Acta2 1:300 ( Abcam #ab5694 ) , anti-Tagln 1:300 ( Abcam ab14106 ) , anti-Pdgfrb 1:500 ( Abcam #ab32570 ) , anti-Tgfbr2 1:100 ( Abcam #ab61213 ) , anti-periostin 1:200 ( Santa Cruz #sc49480 ) , anti-Pdgfra ( R&D Systems #AF1062 ) , anti-vWF 1:250 dilution ( Abcam ab11713 ) , anti-TUBB3 1:1000 dilution ( Abcam ab78078 ) , anti-PLIN1 1:500 dilution ( Abcam ab61682 ) , anti-THY1 1:200 dilution ( Dianova DIA-100-M ) , anti-NES 1:2000 dilution ( Abcam ab134017 ) , anti-LAMP2 1:150 dilution ( Abcam ab37024 ) , anti-PTPRC 1:300 dilution ( Abcam ab ) , and anti-ITGAX 1:100 dilution ( Abcam ab33483 ) . The slides were incubated for 1h at room temperature with Rabbit-on-Rodent AP polymer ( Biocare Medical #RMR625 ) and 10–15 minutes with the substrate Vulcan Fast Red ( Biocare Medical #FR805 ) . Between steps slides were washed 2 x 5 minutes with TBS ( Sigma-Aldrich #T6664 ) . The slides were then dehydrated in an ascending alcohol dilution series and mounted in Cytoseal XYL ( Thermo Scientific #8312–4 ) . Images were taken with a Zeiss Axioplan 2 using the NIS-Elements F v4 . 0 software . EdU staining was employed to quantify in vivo rates of vascular wall cell proliferation in Tcf21iCre/+ , ROSAtdT/+ , Ldlr-/- mice . These mice were weaned onto HFD at 4 weeks of age , administered tamoxifen by a single gavage of 10 μg/40g body weight 1 week before tissue harvest , and sacrificed and evaluated at T0 , 9wks , 12 wks or 15 wks of diet . 5-Ethynyl-2-deoxyuridine ( Santa Cruz , 50μg/g body weight ) was injected intraperitoneal 24 hours before sacrifice . Aortic root tissues were harvested , processed and sectioned , and Tcf21+EdU+ positive cells were quantified and expressed as the % of Tcf21 lineage cells . This percentage was evaluated for significant differences among the various time points evaluated . These data are based on three 40x images at each time point in each animal . Similar experiments were performed to compare the relative rates of cell division in Tcf21lacZ/+ , ApoE-/- compared to ApoE-/- mice . Animals in each group were injected with 1mg of EdU ( Life Technologies , cat#A10044 ) 24 hours before sacrifice and EdU incorporation was imaged according to the manufacturer's instructions with the Click-iT EdU Alexa Fluor 488 Imaging Kit ( Life Technologies ) on OCT ( Tissue-Tek ) frozen sections . Eight sections were assessed from each group with ImageJ software ( NIH ) to determine the ratio of EdU positive cells to the total number of DAPI positive cells within plaque areas . Human atherosclerotic carotid artery lesions were obtained from patients undergoing endarterectomy surgery for stable ( asymptomatic; n = 10 ) or vulnerable ( symptomatic; n = 10 ) carotid stenosis , as part of the Biobank of Karolinska Endarterectomies ( BiKE ) . Samples were collected with informed consent procedure and the study approved by the Local Ethical Committee . Patient demographics , symptom definition and sampling routines were described previously [55] . Micro-dissection laser capture of fibrous cap structures from both subgroups of patient samples was performed using a PALM Microlaser system ( Zeiss , Germany ) according to manufacturer’s instruction . Atherosclerotic lesions , stable and ruptured ( vulnerable ) , were paraffin-embedded , sectioned , and stained with hematoxylin & eosin on RNAse-free glass slides . Sections were pre-treated with UV light at 254nm for 30 minutes to overcome the hydrophobic nature of the membranes and to enhance adhesion of the paraffin-embedded sections . Ten consecutive slides per individual patient were micro-dissected and then pooled for gene expression analysis . In preparation for laser pressure catapulting , sections were first de-paraffinized with Xylene ( 2 x 2 minutes ) and decreasing C2H6O ( Ethanol ) concentrations ( 100% , 96% , 70% each for 1 minute ) . Sections were then rinsed in RNAse-free water , and stained for 10 minutes with Mayer’s hematoxylin ( Sigma , USA ) , again rinsed for 3 minutes in RNAse-free water , and then consecutively stained for 3 minutes with Eosin ( Sigma , USA ) . Finally , samples were dehydrated with increasing concentrations of C2H6O , before briefly being air-dried at room temperature . After micro-dissection and catapulting , the fibrous cap sample was collected into AdhesiveCaps ( Zeiss , Germany ) and 350μl of RLT buffer ( Qiagen , Germany ) was added , and mixed by inversion after closure . The lysate was spun down for 5 minutes at 13 000 rpm and stored at -80°C . RNA extraction from catapulted , micro-dissected samples was performed using the RNeasy Micro Kit ( Qiagen , Germany ) following the manufacturer’s protocol . RNA was quantified by Nanodrop ( Agilent Technologies ) , and RNA quality was verified using an Agilent 2100 Bioanalyzer ( Agilent Technologies ) . Samples required 260/280 ratios >1 . 8 , and sample RNA integrity numbers >9 for inclusion . The TaqMan High Capacity cDNA Transcription Kit was used for cDNA synthesis , and primer assays for TCF21 and 18S ( for housekeeping/endogenous control ) were utilized to detect changes in mRNA expression levels . The mean TCF21 mRNA level ( run in triplicate ) was normalized to the mean 18S mRNA level ( run in triplicate ) to calculate the change in expression for each condition . Results are expressed as means ± standard deviation ( SD ) with number ( n ) of replicates . Statistical comparisons of two groups were performed by two-tailed t-test using Prism if not stated otherwise . Welch’s Correction for unequal variance was applied . | Coronary artery disease ( CAD ) is responsible for the majority of deaths in the Western world , and is due in part to environmental and metabolic factors . However , half of the risk for developing heart disease is genetically predetermined . Genome-wide association studies in human populations have identified over 100 sites in the genome that appear to be associated with CAD , however , the mechanisms by which variation in these regions are responsible for predisposition to CAD remain largely unknown . We have begun to study a gene that contributes to CAD risk , the TCF21 gene . Through genomic studies we show that this gene is involved in processes related to alterations in vascular gene expression , and in particular those related to the smooth muscle cell biology . With cell culture models , we show that TCF21 regulates the differentiation state of this cell type , which is believed critical for vascular disease . Using mouse genetic models of atherosclerotic vascular disease we provide evidence that this gene is expressed in precursor cells that migrate into the disease lesions and contribute to the formation of the fibrous cap that is believed to stabilize these lesions and prevent heart attacks . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [] | 2015 | Coronary Artery Disease Associated Transcription Factor TCF21 Regulates Smooth Muscle Precursor Cells That Contribute to the Fibrous Cap |
The unusual cell cycles of Apicomplexa parasites are remarkably flexible with the ability to complete cytokinesis and karyokinesis coordinately or postpone cytokinesis for several rounds of chromosome replication , and are well recognized . Despite this surprising biology , the molecular machinery required to achieve this flexibility is largely unknown . In this study , we provide comprehensive experimental evidence that apicomplexan parasites utilize multiple Cdk-related kinases ( Crks ) to coordinate cell division . We determined that Toxoplasma gondii encodes seven atypical P- , H- , Y- and L- type cyclins and ten Crks to regulate cellular processes . We generated and analyzed conditional tet-OFF mutants for seven TgCrks and four TgCyclins that are expressed in the tachyzoite stage . These experiments demonstrated that TgCrk1 , TgCrk2 , TgCrk4 and TgCrk6 , were required or essential for tachyzoite growth revealing a remarkable number of Crk factors that are necessary for parasite replication . G1 phase arrest resulted from the loss of cytoplasmic TgCrk2 that interacted with a P-type cyclin demonstrating that an atypical mechanism controls half the T . gondii cell cycle . We showed that T . gondii employs at least three TgCrks to complete mitosis . Novel kinases , TgCrk6 and TgCrk4 were required for spindle function and centrosome duplication , respectively , while TgCrk1 and its partner TgCycL were essential for daughter bud assembly . Intriguingly , mitotic kinases TgCrk4 and TgCrk6 did not interact with any cyclin tested and were instead dynamically expressed during mitosis indicating they may not require a cyclin timing mechanism . Altogether , our findings demonstrate that apicomplexan parasites utilize distinctive and complex mechanisms to coordinate their novel replicative cycles .
Obligate intracellular parasites of the phylum Apicomplexa are responsible for many important diseases in humans and animals , including malaria , toxoplasmosis and cryptosporidiosis . Severity of the disease is tightly linked to parasite burden , and currently , the most successful therapies block parasite proliferation . Apicomplexan parasites use flexible mechanisms to replicate that are different than those of their hosts . In endodyogeny each round of duplication is completed with assembly of two internal daughters [1] . Alternatively , multiple buds can be formed inside of the mother or emerge from its surface in the processes called , endopolygeny and schizogony , respectively [2] . Toxoplasma gondii undergoes endodyogeny in intermediate host stages , but replicates by endopolygeny in the definitive feline host . Fundamental differences between division modes are embedded in the features of the apicomplexan cell cycle comprised of two chromosome cycles [2 , 3] . During nuclear cycles , chromosomes are replicated and segregated without budding , while each round of chromosome replication in the budding cycle leads to production of the daughter parasites . Ultimately , the number of nuclear cycles determines the scale of the parasite progeny . We recently showed that the ability to switch between chromosome cycles is partially linked to the unique bipartite structure of the T . gondii centrosome [3] . Weakening or separation of the outer centrosomal core that controls budding favors the nuclear cycle , while the strong association of the outer core with the inner core promotes cytokinesis and the budding cycle [3] . Endodyogeny of T . gondii tachyzoites represents one the simplest modes of replication and important cell cycle transition points where potential checkpoints may operate have been defined [2 , 4–19] . The G1 phase of T . gondii endodyogeny comprises half of the division cycle , and like other eukaryotes , canonical housekeeping tasks preparing for S phase commitment are performed in the apicomplexan G1 period [4 , 9 , 13 , 15 , 18 , 19] . Evidence also indicates that cell cycle exit to form dormant developmental stages as well as drug-induced dormancy is controlled by mechanisms acting at the transition from M/C into early G1 [4 , 20 , 21] . Duplication of the centrosome marks the transition from G1 to S phase [3 , 6 , 15 , 17] , and we have defined some of the components of this critical transition in T . gondii that should be targets of a G1/S checkpoint mechanism [17] . A peculiar feature of apicomplexan replication is the short ( or absent ) G2 phase [1 , 2] that is thought to be marked by natural S phase populations that possess partially duplicated genomes ( 1 . 7–1 . 8N DNA ) [1 , 13] . Resolving the molecular basis for this important transition should help solve the mystery of these unusual DNA distributions . During the apicomplexan mitosis numerous specialized structures are replicated , built or converted in precise order to produce healthy infectious daughters ( for review see [2 , 22] ) . Our previous work and the studies of others have established that duplication of the bipartite centrosome is coordinated with division of the centrocone that holds the mitotic spindle [3 , 6 , 10 , 23] , and also with the replication and segregation of the bundled centromeres [23 , 24] , which in turn , is synchronized with assembly of the basal and apical complexes of the future daughter [25–28] . The temporal-spatial coordination of these overlapping processes likely requires similarly complex regulatory machinery , for which the molecular basis is still largely unknown . In eukaryotes , cell cycle progression is governed by the activity of cyclin-dependent kinases ( Cdks ) and their regulatory cofactors , cyclins [29 , 30]; dynamic expression of the latter provides clockwise control of Cdk function . Cdk4/6-cyclin D complexes support the progression of G1 phase , and Cdk2 complexes with cyclins E , A and B govern the progression and fidelity of DNA replication in S phase and chromosome segregation in mitosis [31] . Cdks functions were originally thought to be restricted to cell cycle regulation , however , today we understand that activated Cdk-cyclin complexes are master regulators of such major biological processes as transcription , RNA processing , translation and development [30] . Extrapolating current models of cell cycle checkpoints that involve Cdk-cyclins to eukaryotes in general is challenging , as there are many branches of the eukaryotic tree where cell division is quite unusual and the molecular controls are likely to be different [29] . This includes the large group of obligate intracellular parasites from the phylum Apicomplexa . Mining the initial genomes of important disease causing apicomplexans has revealed major differences [1 , 32–34] characterized by the reduction of components and also the complete absence of the key regulatory elements , including canonical cyclins [35] , major cell cycle Cdks [33 , 34] and their immediate downstream effectors [1] . The lack of conserved cell cycle factors of higher eukaryotes indicates there are significant changes in the cell cycle molecular machinery of these ancient protozoa . Here we describe the first comprehensive study of Cdk-related kinases ( Crks ) and cyclins in T . gondii . Using genetic approaches , we have analyzed the function of seven TgCrks and four TgCyclins and uncovered major cell cycle TgCrks controlling the replication of the tachyzoite stage of T . gondii . Our results demonstrate that , unlike the traditional eukaryotic cell cycle , the intricate division of apicomplexan parasites is regulated by multiple essential Crks acting independently at several critical transitions and in unusual spatial contexts .
To determine the core Cdk-cyclin complexes that regulate division in T . gondii , we systematically searched the parasite genome and identified ten genes encoding a kinase domain that included a cyclin-binding sequence ( C-helix ) ( S1 Fig ) [30] . Following standard convention , we named these factors Cdk-related kinases ( Crks ) until cyclin-dependent activation of the kinase can be established [36 , 37] . T . gondii Crks ( TgCrks ) were a diverse group of proteins ranging from 34 to 212kDa ( S1A Fig ) that also varied in mRNA abundance ( S1B Fig ) . T . gondii transcriptome data ( ToxoDB ) indicated that eight TgCrks were expressed in tachyzoites and/or bradyzoites , and mRNA profiles for two kinases , TgCrk2-L1 and TgCrk5-L1 indicated they were restricted to the definitive or environmental life cycle stages; merozoite and/or sporozoite ( ToxoDB ) . The profiles of TgCrk4 , TgCrk5 and TgCrk6 mRNAs were cyclical in tachyzoites ( S1B Fig ) [4] , which was confirmed at the protein level for TgCrk4HA and TgCrk6HA by endogenous epitope tagging ( S3 Fig ) . The dynamic cell cycle regulation of the three TgCrk factors differs significantly from the typical constitutive expression of Cdks of studied model eukaryotes [38 , 39] . Phylogenetic analysis of TgCrks defined the general Cdk families present in the Apicomplexa phylum [34] . Putative Crks from T . gondii , P . falciparum [33 , 34] , Theileria annulata and Cryptosporidium parvum were compared to the ancestral free-living unicellular eukaryote Chromera velia , and the extensively studied Cdks of human cells . The analysis sorted ten TgCrks into eight general phylogenetic clades with five groups restricted to the superphylum Alveolata ( Fig 1 , pink shade ) : TgCrk5 and TgCrk5-L1 were apicomplexan adaptations . Kinases in three other clades were shared with a recognizable higher eukaryotic counterpart that included several kinases known to regulate cell cycle and gene expression . Specifically , TgCrk1 grouped with the Cdk11 family kinases that regulate mRNA synthesis and maturation ( Fig 1 , green shade ) , while TgCrk7 was similar to the Cdk-activating kinase ( CAK ) , HmCdk7 ( Fig 1 , yellow shade ) . Lastly , TgCrk2 clustered with the family of eukaryotic cell cycle regulators , in particular , with neuronal HmCdk5 ( Fig 1 , light blue shade ) . To define the function of TgCrks , we constructed tet-OFF conditional knockdown mutants by replacing the native promoter with a tetracycline-regulatable promoter in the Tati-RHΔku80 strain [24 , 40] . Each kinase was concurrently tagged with a 3xHA-epitope fused to the N-terminus ( S1D Fig , diagram ) . We successfully generated tet-OFF mutants for seven TgCrks demonstrating that promoter replacement and N-terminal HA-tagging were remarkably tolerated in the tachyzoites ( Fig 2 and S1F Fig ) . The only other TgCrk factor expressed in tachyzoites is TgCrk5 , which is the subject of another project and was not studied here ( Naumov and White , personal communication ) . Immunofluorescence assays ( IFA ) of HATgCrks determined that most of these kinases were predominantly nuclear in tachyzoites ( Fig 2 , -ATc conditions ) with the exception of HATgCrk2 , which was expressed throughout the cell , and HATgCrk4 , which was exclusively localized to the cytoplasm . Despite sharing the same tet-OFF promoter , individual HATgCrks showed a wide range of protein abundance that closely matched the expression level of the factors regulated by their native promoters , indicating there are major post-transcriptional mechanisms controlling TgCrk levels in tachyzoites ( S1E Fig ) . The HATgCrk proteins were all successfully down regulated by a 24 h incubation with 1μg/ml anhydrotetracycline ( ATc ) ( Fig 2 , IFA and Western blot analysis ) . Exploiting the ATc-induced conditional knockdown , we determined by plaque growth assay ( Fig 2 ) that four TgCrks were essential , two were required for tachyzoite growth and only TgCrk8 was dispensable . Quantitative growth rates ( 24 h ) for the TgCrk4 and TgCrk6 tet-OFF mutants with or without ATc treatment confirmed that TgCrk4 was required and TgCrk6 was essential for tachyzoite growth ( S1F Fig ) . IFA analysis determined that ATc-induced growth arrest of TgCrk3 and TgCrk7 tet-OFF mutants was cell cycle independent ( Fig 2 and S2B Fig ) consistent with the potential role in regulating general cellular processes required around the entire cell cycle . Recent studies of TgCrk7 , and its P . falciparum ortholog Pfmrk , implicated a role in transcriptional regulation [41 , 42] and PfCRK3 kinase complexes were associated with chromatin-dependent regulation of gene expression [37] . Based on parasite morphology , the loss of TgCrk2 appeared to block tachyzoites in the G1 phase and knockdown of TgCrk1 , TgCrk4 or TgCrk6 resulted in extensive mitotic and cytoskeletal defects , consistent with growth arrest in the S/M/C half of the cell cycle ( S2A Fig ) . These four kinases were selected for further characterization in this study . In general cyclins are poorly conserved , although the presence of a cyclin box and one or more destruction motifs permits these genes to be identified by genome mining . Utilizing this approach we identified seven novel cyclin factors in the T . gondii genome ( S4A Fig ) [35] . Only cyclins related to P- , H- , L- and Y- types were found , while no canonical A- , B- , D- , and E- types , that are vital to higher eukaryotic cell division , were identified [43] . Based on the extensive transcriptome data ( see ToxoDB ) , five of the seven TgCyclins appeared to be expressed in tachyzoites ( log2 RMA value higher then 6 , S4B Fig ) . To determine expression and localization of TgCyclins in tachyzoites , we epitope tagged the C-terminus of TgCycH , TgCycL , TgCycY with a 3xHA-epitope by genetic knock-in . IFA analysis showed that TgCycHHA , TgCycLHA and TgCycYHA , were moderately expressed and localized to the nucleus in tachyzoites ( Fig 3A ) , and remarkably , TgCycYHA was the only oscillating cyclin with peak expression in the G1 phase ( Fig 3B ) . To visualize the lower abundant cyclin , TgPHO80 , we engineered transgenic parasites ectopically expressing TgPHO80 that was N-terminally tagged with a 3xmyc epitope fused to a FKBP destabilization domain ( DDmyc ) , which permits conditional expression using the small molecule Shield 1 [15 , 44] . After 3 h stabilization with Shield 1 ( 100nM ) , DDmycTgPHO80 was found in large cytoplasmic speckles ( Fig 3A ) , which was the only cyclin exclusively localized to the tachyzoite cytoplasm . Utilizing a similar knockdown approach as was used for the TgCrks ( S1D Fig ) , we successfully converted by genetic knock-in four TgCyclin genes to tet-OFF mutant alleles: TgPHO80 , TgCycH , TgCycL and TgCycY ( Fig 3C , Western Blot analysis ) . ATc-knockdown of TgPHO80 slowed the rate of replication ( S4C Fig ) and increased the fraction of parasites with a single centrosome ( 76 . 5% ± 9 . 8 vacuoles after 48 h with 1μg/ml ATc compared to 49% ± 1 vacuoles in -ATc conditions ) indicating the G1 phase was lengthened by the loss of TgPHO80 . The ATc-induced depletion of nuclear cyclins , TgCycL and TgCycH , caused tachyzoite growth arrest . TgCycL deficiency led to mitotic death ( S2A Fig ) and knockdown of TgCycH resulted in a quick non-specific growth arrest ( Fig 3C and S2B Fig ) . The conditional knockdown of TgCycY with ATc demonstrated this cyclin was not essential for tachyzoite growth ( Fig 3C and growth quantification in S4C Fig ) . The conditional knockdown of the tet-OFF TgCrk2 mutant appeared to cause a cell cycle arrest in G1 , which we confirmed by analyzing centrosome duplication . As expected , the majority ( ~80% ) of TgCrk2-depleted parasites ( +ATc ) possessed a single centrosome ( Fig 4A ) indicative of G1 phase arrest . Since canonical G1 cyclins ( e . g . D- and E- type ) [29] were absent from the T . gondii genome , it was of particular interest to identify the T . gondii cyclin that binds TgCrk2 . We genetically engineered a series of dual epitope-tagged strains by first knocking-in 3xHA epitope tag into the TgCrk2 gene followed by stable ectopic expression of four different TgCyclins that were epitope tagged with 3xmyc ( see S1 Table for list of transgenic strains and Material and methods for production details ) . TgCrk2 protein complexes were affinity purified with antibody against the epitope tag and then probed with alternative antibody to define the interacting cyclin . We detected a weak interaction between TgCrk2 and TgCycH ( Fig 4B ) confirming previous yeast two-hybrid screens [45] , however , TgCrk2HA formed the most abundant complexes with cyclin TgPHO80 ( Fig 4B ) , which corroborates the G1 phenotype we observed for the TgPHO80 tet-OFF mutant following ATc-knockdown ( Fig 3C and S4C Fig ) . Conditional loss of TgCrk1 in the tet-OFF mutant resulted in severe chromosome mis-segregation and accumulation of deformed zoites ( Fig 2 ) . To more precisely define the TgCrk1-loss phenotype , we focused on the early mitotic steps just prior to daughter bud formation . The MORN1 protein associates with two compartments that are duplicated in mitosis , the spindle compartment ( centrocone ) and basal complex referred to here as the MORN-ring ( Fig 5A , image a ) [27 , 28] . Utilizing the MORN1 marker , we determined that in parasites lacking TgCrk1 ( +ATc for 16 h ) the development of the daughter MORN1 rings was defective ( Fig 5A and 5B , compare image a to c , e ) . Deficiency of the basal complex became more evident when newly produced alveolar sacs ( Fig 5A and 5B , image f ) accumulated near fragmented MORN-rings in the late stages of mitosis ( Fig 5A and 5B image e ) . In addition , the normal coordination of mitosis with cytokinesis failed and the IMC compartment appeared as an unstructured mass ( Fig 5A and 5B , image h ) . In addition , TgCrk1 deficiency affected the structural integrity of the tachyzoite apical end ( Fig 5C ) . Following ATc treatment the robust cone-shaped apical cap ( ISP1 staining ) [46] became a weak rod-like or lopsided structure that caps a deformed IMC1-positive mass ( Fig 5C ) . Despite severe defects in mitosis caused by the loss of TgCrk1 , the duplication and segregation of centromeres , centrosomes and the plastid ( S5A Fig ) and nuclear division were evident ( Fig 5A , image i ) . Altogether these results indicate that TgCrk1 primarily regulates cytokinesis but not karyokinesis . Constitutively expressed TgCrk1Ty ( S5B Fig ) and TgCycLHA are similarly localized and knockdown phenotypes of TgCycL and TgCrk1 tet-OFF mutant parasites were also similar ( Figs 2 and 3 and S2A Fig ) , indicating these two proteins may be paired in T . gondii . Indeed , co-IP experiments using the dual epitope-tagging approach employed with TgCrk2 above , confirmed preferential interaction of TgCrk1 with TgCycL ( Fig 5D ) . Furthermore , IFA of the dual-tagged strain co-expressing TgCrk1HA and TgCycLmyc revealed tight co-localization of these factors in a unique sub-nuclear compartment ( Fig 5E ) that was independent of the nucleolus ( TgNF3 , [47] ) , centromere-compartment ( TgCenH3 , [23] ) , centrocone ( MORN1 , [26] ) , or nascent particles containing the RNA polymerase complex ( TgRPB4 , [48] ) ( S5C Fig ) . Conditional knockdown of TgCrk4 and TgCrk6 tet-OFF mutants also caused severe mitotic defects ( Fig 2 and S2A Fig ) indicating that similar to TgCrk1 , these Apicomplexa-specific kinases function in the second half of the tachyzoite cell cycle . IFA analysis of endogenously tagged TgCrk6HA and TgCrk4HA determined that these proteins have different subcellular localization ( Fig 2 ) . TgCrk4HA was distributed in large cytoplasmic aggregates with accumulation in the apical perinuclear region ( S3C Fig ) , while TgCrk6HA extended its nuclear localization to the centromeric region ( S3D Fig , CenH3 and Centrin1 staining ) [23] . To build clues to TgCrk6 function , we performed detailed IFA analysis following short term ATc treatment ( 16 h ) using the MORN1 ( mitotic structures ) and IMC1 ( cytoskeleton ) cell cycle markers . The normal assembly of the daughter scaffold is initiated in late S phase followed by duplication/separation of the centrocone ( Fig 6A , -ATc ) [2 , 3] . Cytokinesis progressed in TgCrk6-deficient parasites , yet the centrocone spindle compartment was not properly duplicated as evidenced by the single MORN1-positive dot positioned between two growing daughter buds ( Fig 6A , +ATc ) . Other evidence supported TgCrk6 function in karyokinetic processes . ATc-downregulation of TgCrk6 disrupted the usual dynamics of kinetochores visualized by co-staining of the kinetochore complex component , TgNdc80myc , and acetylated Tubulin A that labels active sites of the microtubule assembly including spindle and internal daughters ( Fig 6B ) [24] . In normal parasites , the TgNdc80myc signal largely disappeared at mid-bud development ( Fig 6B , -ATc image e ) . By contrast , TgCrk6 deficient parasites ( +ATc ) retained single assembled kinetochores well into the budding process ( Fig 6B , image g ) . Longer term ATc incubations ( >24 h ) amplified the loss of coordination between cytokinesis and karyokinesis leading to the catastrophic phenotype of severe DNA mis-segregation and assembly of buds lacking DNA shown in Fig 2 . These results combined with the observed accumulation of the nuclear TgCrk6HA in the centromeric region during peak expression in S/M phase ( S3D Fig ) supported a key role for TgCrk6 in T . gondii spindle regulation . Similar deficiency to split a spindle was recently described in the knockdown mutant of the distantly related P . falciparum CRK4 ( Fig 1 ) [49] . Similar to TgCrk6 , knockdown of TgCrk4 caused defects in mitosis ( Fig 2 ) , although TgCrk4 differed from TgCrk6 in being localized to the cytoplasm rather than the nucleus ( S3B versus S3D Fig ) . This difference led us to examine the role of TgCrk4 in regulating the cytoplasmic components of the mitotic machinery . Asexual stages of T . gondii divide by enclosed mitosis ( as do most apicomplexans ) that coordinates attachment of nuclear centromeres to kinetochores/spindle and to a unique centrosome containing two independent functioning core structures [3 , 22] . Consistent with a role in controlling mitosis through cytoplasmic structures , down regulation of TgCrk4 with 1μg/ml ATc led to defective duplication of both centrosomal cores ( Centrin1/outer core and CEP250myc/inner core ) , but did not affect centromere duplication/segregation ( CenH3 marker ) or nuclear division ( Fig 7A ) . Interesting , plastid segregation , which is controlled by the centrosome [50] , was also defective in parasites lacking TgCrk4 ( Fig 7A , TgAtrx1 marker ) . Although TgCrk4-deficient parasites showed abnormal centrosome replication ( under and over reduplication ) ( Fig 7A and 7B ) , we did not observe uncoupling of the centrosome cores ( Fig 7C ) as we have documented in some temperature sensitive mutants [3] . Moreover , we found that centrosome re-duplication occurred around assembled kinetochores ( Fig 7D ) despite the disruption in normal centrosome stoichiometry . The proper ratios of centrosome to kinetochore observed in a regular tachyzoite mitosis are established ( Fig 7D , -ATc; 2:1 images a , b; 2:2 images c , d; 2:0 images e , f ) . By contrast , down regulation of TgCrk4 led to abnormal stoichiometry of 4 centrosomes to 2 assembled kinetochores ( Fig 7D , +ATc , images g , h ) . Further examination revealed that only one of the reduplicated centrosomes remained associated with the nucleus ( Fig 7C , inset ) , which may explain why the loss of TgCrk4 did not lead to unregulated karyokinesis . Altogether , our results support the role of TgCrk4 kinase in the regulation of centrosome duplication and segregation during mitosis within the context of other essential mitotic regulatory controls such as TgCrk6 above . Interestingly , TgCrk4 and TgCrk6 were cyclically expressed during tachyzoite replication ( S3 Fig ) with the peak expression in S/M phase consistent with functions in regulating mitotic processes . In higher eukaryotes , mitotic Cdk activity is typically controlled by an oscillating cyclin partner , while the Cdk protein is constitutive [29 , 38 , 39] . To identify the cyclin partners for TgCrk4 and TgCrk6 , we performed co-IPs from dual tagged strains expressing TgCrk4HA or TgCrk6HA and four different ectopically expressed TgCyclins ( See Material and methods ) . No detectable interaction between the TgCyclins tested and TgCrk4 and TgCrk6 was observed ( S4D Fig ) suggesting T . gondii may have become dependent on direct mechanisms of dynamic expression to regulate mitotic TgCrk4 and TgCrk6 factors leading to the loss of a periodic activating cyclin partner .
The molecular basis of cell cycle regulation in eukaryotes has been mainly shaped by studies in one branch of eukaryotes , Unikonta that includes the clades animalia , fungi and amoebas [29] . This is a significant deficiency because the replication biology of eukaryotes from the Bikonta branch , comprised of the three supergroups , the Excavata , SAR ( Stramenopiles , Alveolates and Rhizaria ) and Archaeplastida , is quite extraordinary , if also beyond our reach experimentally [51–53] . From this point of view , our genetic analysis of T . gondii Cdk-related kinases and cyclins provides much needed insight into the cell cycle regulation of an ancient protozoan from the SAR supergroup . One of the core findings of this study is the surprising complexity and unusual regulation of cell cycle controls that are essential for Apicomplexa cell division . The results of multiple gene knockouts in higher eukaryotes reveals that a single active Cdk ( Cdk1/2 family ) is sufficient to sustain basic chromosome segregation in the somatic cells of both multicellular and unicellular eukaryotes [29 , 54] . By contrast , this study and a second project in progress ( TgCrk5 studies , Naumov and White , personal communication ) have established that T . gondii requires five Crks to successfully regulate the peculiar parasite cell cycle called endodyogeny ( Fig 8 ) . Two Crks regulate centrosome duplication ( TgCrk4 ) and organization of the daughter bud cytoskeleton ( TgCrk1 ) during interwoven S , M and C phases . T . gondii has also evolved independent controls for the restriction or START checkpoint in G1 ( TgCrk2 ) , the DNA licensing checkpoint in S phase ( TgCrk5 , Naumov and White , personal communication ) and the spindle assembly checkpoint ( TgCrk6 ) acting at the metaphase to anaphase transition in mitosis ( Fig 8 ) . A Recently published study of P . falciparum CRK4 confirms that cell cycle checkpoints are regulated by the related kinases across Apicomplexa phylum [49] . Similar to TgCrk6 , nuclear PfCRK4 was dynamically expressed in S/M phase ( late trophozoite to schizont during blood stage ) and upon downregulation parasites lost the ability to split the spindle and , consequently properly segregate chromosomes ( Fig 2b , f and g in [49] ) . The high number of the putative cell cycle checkpoints was unexpected and this favors a model of much tighter cell cycle regulation in the Apicomplexa than previously thought [1] . In fact , the lack of reversible and abundant catastrophic phenotypes observed in T . gondii cell cycle mutants [9] , which was often interpreted as a lack of cell cycle controls , is likely a consequence of the complexity of this system . We propose that apicomplexan parasites evolved separate Crks for individual cell cycle stages to facilitate switches between flexible division modes [2 , 3] . During the chromosome cycle of schizogony or endopolygeny , the G1 phase is completed only once and is uncoupled from the multiple rounds of the S/M phase , which are , in turn , uncoupled from the budding process until the very last unified S/M/C phase . Therefore , evolution of multiple Crks offers independent control of the segments , permitting modular regulation of the complex cell cycles , yet , leaving an open question of the master regulator ( s ) of the switch . Analyzing the T . gondii cell cycle we noticed multiple parallels in topology of the cell cycle regulation between apicomplexans and a few studied Bikonta models , particularly , plants . First , similar to plants , mitosis in apicomplexans is regulated by clade-specific Cdks ( Fig 1 ) [49 , 55] . Second , mitotic TgCrk4 , TgCrk5 ( Naumov and White , personal communication ) , TgCrk6 and PfCRK4 [49] are dynamically expressed ( protein and mRNA ) , which is a distinctive feature of the CDKB family kinases in plants [55 , 56] . Third , similar to most Archaeplastida members , apicomplexan parasites do not possess or encode a highly diverged Cdc25 phosphatase ortholog ( EupathDB ) [56 , 57] , which primary function is to activate mitotic Cdks . Interestingly , based on the functional parallels between Arabidopsis thaliana CDKB1;1 and Drosophila melanogaster Cdc25 , plant biologists proposed that the Cdc25-controlled onset of mitosis may have been evolutionary replaced by plant-specific B-type Cdk pathway [57] . Whether the similar scenario had taken place in Apicomplexa evolution will require further studies , involving broader analysis of the Bikonta organisms . Unfortunately , the current limitation of the bikont studies also does not permit us to make a definitive conclusion of whether the plant-like features of the Apicomplexan cell cycle were inherited at the time of the Unikonta and Bikonta diversion or were the result of a secondary symbiosis of the Chromoalveolata and red algae [58] . The majority of characterized cell cycle cyclin-Cdk complexes are composed of constitutive and dynamic subunits with the cyclin generally the oscillating partner [38] . Consistent with the concept that functional topology is more important than the conservation of individual parts [59] , we found that T . gondii cyclins differed from their high eukaryotic counterparts in being pre-dominantly constitutively expressed . Intriguingly , a single oscillating cyclin , TgCycY , was not essential for tachyzoite division nor did it interact with the seven TgCrks we analyzed ( S4D Fig ) and , therefore , a possible Cdk-independent role will need to be explored in future studies [60] . In fact , only two cell cycle complexes , TgCrk2-TgPHO80 and TgCrk1-TgCycL , were detected in which both subunits seem to be constitutively expressed in tachyzoites , suggesting mechanisms other then cyclin-binding that regulate cell cycle activity of TgCrk1 and TgCrk2 . On the contrary , mitotic TgCrk4 and TgCrk6 are rare examples where Cdk-related kinases are dynamically expressed and not found in the complex with TgCyclins ( S4D Fig ) . Given that the role of cyclin is to provide a temporal context to Cdk function , it is possible that T . gondii no longer needs cycling partners for TgCrk4 and TgCrk6 . There is reason to speculate that constitutively expressing mitotic kinases might not be ideal given the complexity of the apicomplexan mitosis that is associated with the extensive de novo biosynthesis of the motility and invasion apparatus of daughter parasites [4] . Delivering the master conductors "just-in-time" would avoid accidentally triggering the cascade of mechanisms that unfold during mitosis and cytokinesis before the parasite is ready to egress . This hypothesis , however , cannot explain the absence of cyclin interaction with the constitutively expressed TgCrk3 and TgCrk8 ( S4D Fig ) . Interestingly , our study revealed that the only T . gondii Crks that have orthologs in other eukaryotes ( Fig 1 , TgCrk1 , TgCrk2 and TgCrk7 ) interacted with conventional cyclins that were also expressed in other eukaryotes ( TgCycL , TgPHO80 and TgCycH ) . Since all the novel TgCrks were orphan , it is tempting to suggest that a non-cyclin factor may have co-evolved that acts as an oscillating component in the complexes with novel TgCrks . Alternatively , these unusual TgCrks may function without a cyclin partner . Future unbiased approaches will be needed to sort out these possibilities , and identify if there are other protein co-factors that have replaced the traditional cyclin partner . Another core finding was the strong molecular support for the remarkable physical partitioning of Apicomplexan cell division functions . Similar to other apicomplexans , T . gondii divides by enclosed mitosis where a set of tethered structures localized in the nucleus or in the cytoplasm must be constructed/deconstructed ( kinetochores , spindle microtubules and striated fibers ) , duplicated/segregated ( bipartite centrosome and centrocone ) or assembled ( buds ) in a timely manner to produce infectious progeny [2 , 3 , 8] . Moreover , in our previous study [3] , we demonstrated that T . gondii has divided the regulatory responsibility for karyokinesis and cytokinesis between two unique centrosome cores that have fixed orientation to nuclear and cytoplasmic biosynthetic events . Results of our study here largely support this nuclear/cytoplasmic organization , which was likely needed to overcome limitations of enclosed mitosis . T . gondii has evolved the nuclear TgCrk6 mechanism to control events requiring the intranuclear spindle , while cytoplasmic TgCrk4 regulates centrosome duplication and associated plastid segregation . While proper assembly of the nuclear spindle and cytoplasmic daughter bud were essential processes , reduplication of the centrosome in TgCrk4-deficient parasites only partially affected tachyzoite survival . We believe that apicomplexans may have relaxed the control of centrosome reduplication in order to more easily adapt cell division to the scale needed in different hosts [3 , 22] . It should be noted that dissolution of the nuclear membrane during open mitosis in higher eukaryotes is a key difference with apicomplexan cell division that may permit extensive re-arrangements of the nuclear ( e . g . kinetochore , spindle ) and cytoplasmic ( e . g . centrosome ) structures by a single Cdk . A possible exception to the nuclear/cytoplasmic functional organization is the assembly of the cytoplasmic daughter buds that was unexpectedly controlled by the nuclear TgCrk1-TgCycL complex . Initiation of the daughter buds near the centrocone , spindle pole , which continues to grow with the progression of mitosis , is a specialized event occurring in the budding cycle of Apicomplexa [2 , 25] . How does the nuclear TgCrk1 regulate assembly of a cytoplasmic structure ? Recent studies of eukaryotic splicing kinase Cdk11 discovered an unexpected role for this kinase in mitotic progression [61] . It has been shown that activity of Cdk11 is required to regulate sister chromatid cohesion [62] . Since knockdown of T . gondii TgCrk1 did not affect DNA segregation , it is possible that the role of TgCrk1 was re-adapted to regulate splicing of mRNAs whose products will be required to control assembly of the daughter buds . Our hypothesis is supported by the fact that many components of the cytoskeleton are delivered “just-in-time” during cell cycle progression [4] . In future studies , the analysis of transcriptome changes caused by TgCrk1 deficiency will help determine whether this kinase operates primarily as a regulator of mRNA expression . Transmission stages are formed at the end of each apicomplexan life cycle , that are , in most species , specialized G1/G0 states . For example , mature bradyzoites and sporozoites of T . gondii remain growth arrested until appropriate external signals from the same or new host trigger recrudescence or de-differentiation , respectively , resulting in re-entry into G1 phase of the asexual proliferative cycle ( Fig 8 ) . In higher eukaryotes , Cdk4/6-Cyclin D complexes are responsible for the cell fate decision to divide or differentiate ( restriction checkpoint ) [63 , 64] , which are factors not present in the Apicomplexa . Here we showed that T . gondii parasites have replaced the canonical G1 machinery of higher eukaryotes with a novel complex of TgCrk2 ( Cdk5 family ) and TgPHO80 cyclin to regulate progression through the tachyzoite G1 phase . Recent studies in kinetoplastids and another apicomplexan , Plasmodium berghei , determined that a related G1 Crk and a P-type cyclin regulate developmental stages in these protozoans , which is similar to our discoveries of TgCrk2 function in T . gondii ( Fig 8 ) [35 , 65 , 66] . It is also worth noting that T . gondii possesses paralogs of TgCrk2 and TgPHO80 cyclin that are developmentally regulated ( ToxoDB , S1 and S4 Figs ) opening the possibility that the TgCrk2-TgPHO80 pathway has diverse functions in development and could also be involved in the regulation of drug-induced dormancy . In conclusion , the systematic approach we have used to analyze the cell cycle machinery in T . gondii has opened the way into learning how cell division is regulated in apicomplexan parasites . Our study has also evoked important issues that still need to be addressed . For example , greater numbers of TgCrks requires greater coordination; so is there a master regulator after all ? Can the complexity of mitosis coupled to cytokinesis in T . gondii division explain a rise of multiple mitotic Crks ? What are the mechanisms responsible for periodic Crks , and how are these mechanisms controlled ? Does the lessons learned here translate to the exotic mitotic mechanisms in related alveolates [51 , 52] ? Clearly , there is still much to be done to understand the molecular basis of apicomplexan cell division , fortunately , we now have important new genetic tools to go forward .
No human subjects or animal research were used in this study . T . gondii strains RHΔhxgprt [67] , Tati-RHΔku80 [40] , and RHΔku80Δhxgprt [68] were cultured in human foreskin fibroblasts ( HFF , gift Dr . David Roos , University of Pennsylvania ) according to published protocols [69] . Viability of transgenic strains was measured in plaque assays as previously described [15] . Monolayers of HFF cells were infected with 150–200 parasites per 35mm dish and individual plaques formed after 6 days were stained with crystal violate and counted . To determine division rates , parasites were counted by IFA using α-IMC1 ( surface; kindly provided by Dr . Gary Ward , University of Vermont ) antibody and DAPI ( nucleus ) in 50 randomly selected vacuoles in three biological replicates after 24 hours growth . Statistical significance was calculated using an unpaired T-test and Bonferroni correction ( Prism6 ) . Transgenic strains and primers created in the study are listed in S1 Table . Monolayers of HFF cells were grown on coverslips and infected with parasites under indicated conditions . Cells were fixed in 4% PFA , permeabilized with 0 . 25% Triton X-100 , blocked in 1%BSA and incubated sequentially with primary and secondary antibody [16] . The following primary antibodies were used: mouse monoclonal α-ISP1 ( clone 7E8 ) [46] and α-Atrx1 ( clone 11G8 ) ( kindly provided by Dr . Peter Bradley , UCLA ) [75] , α-TgCenH3 [23] ( kindly provided by Dr . Boris Striepen , University of Georgia , Athens ) , α-acetylated alpha Tubulin ( Abcam ) , rat monoclonal α-HA ( clone 3F10 , Roche Applied Sciences ) , rabbit polyclonal α-myc ( Cell Signaling Technology ) , α-Human Centrin 2 [17] , α-MORN1 ( kindly provided by Dr . Marc-Jan Gubbels , Boston College ) [27] and α-IMC1 ( kindly provided by Dr . Gary Ward , University of Vermont ) . Alexa-conjugated secondary antibodies of the different emission wavelengths ( Molecular Probes , Thermo Fisher Scientific ) were used at a dilution of 1:1000 . Stained parasites on the coverslips were mounted with Aqua-mount ( Lerner Laboratories ) , dried overnight at 4°C , and viewed on a Zeiss Axiovert Microscope equipped with 100x objective . Images were collected and processed using Zeiss Zen software and were further processed in Adobe Photoshop CC using linear adjustment when needed . Transgenic parasites co-expressing epitope-tagged TgCrks and TgCyclins were grown for 30–32 hours at 37°C . When expression of the factor was conditional , 100nM Shield1 was added for the last 3 hours of incubation . Parasites ( 3x108 ) were collected , washed in PBS and lysed in 1xPBS with 0 . 5% NP-40 , 400mM NaCl , protease and phosphatase inhibitors ( Thermo Fisher Scientific ) on ice for 30 min . Total protein extract obtained by centrifugation at 21 , 000xg , 10 min , 4°C was divided and incubated with α-HA or α-myc magnetic beads ( MBL International ) for 1 h at room temperature . Isolated protein complexes on the beads were washed three times with the lysis buffer and eluted by heating in the Leamlli sample buffer at 65°C , 10 min . Protein extract before and after pull-down , and purified protein complexes were analyzed by Western blotting . To prepare samples of the total extracts , parasites were purified by filtering through 3μm polycarbonate filters ( EMD Millipore ) , washed in PBS , re-suspended with Leammli loading dye and lysed at 65°C for 10 min . To analyze individual fractions after immunoprecipitation , an aliquot of the fraction was mixed with Leammli loading dye and heated for 10 min at 65°C . After separation on SDS-PAGE gels , proteins were transferred onto a nitrocellulose membrane and probed with monoclonal α-HA ( 3F10 , Roche Applied Sciences ) , α-myc ( Cell Signaling Technology ) and α-Tubulin A ( 12G10 , kindly provided by Dr . Jacek Gaertig , University of Georgia ) antibodies . After incubation with secondary HRP-conjugated antibodies , proteins were visualized by enhanced chemiluminescence detection ( PerkinElmer ) . Because of the lack of cell cycle gene annotation in apicomplexans , we searched the T . gondii genome ( toxoDB ) for protein kinases with cyclin binding C-helix and proteins containing a cyclin box showing similarity to mammalian Cdks and cyclins , respectively . Then , to reduce complexity of the analysis we first identified Cdk classes preserved in apicomplexans by comparing Cdk-related kinases of T . gondii and Cdks of human cells ( ncbi . org ) . The evolutionary history of T . gondii Crks was inferred by using the Maximum Likelihood method based on the Whelan And Goldman model [76] and were conducted in MEGA7 [77] . The analysis involved 54 amino acid sequences of the putative Cdk-related kinases from T . gondii ( Tg , ToxoDB ) , P . falciparum ( Pf , PlasmoDB ) , T . annulata ( TA , PiroplasmaDB ) , C . parvum ( Cgd , CryptoDB ) , C . velia ( Cvel , CryptoDB ) and pre-selected Cdks from human cells ( Hs , ncbi . org ) that are listed in S1 Table . All positions containing gaps and missing data were eliminated . There were a total of 158 positions in the final dataset . The bootstrap consensus tree inferred from 100 replicates . | Apicomplexan parasites are unicellular eukaryotes that replicate in unusual ways different from their multicellular hosts . From a single infection , different apicomplexans can produce as few as two or up to many hundreds of progeny . How these flexible division cycles are regulated is poorly understood . In the current study we have defined the major mechanisms controlling the growth of the Toxoplasma gondii acute pathogenic stage called the tachyzoite . We show that T . gondii tachyzoites require not only multiple protein kinases to coordinate chromosome replication and the assembly of new daughter parasites , but also each kinase has unique responsibilities . By contrast , the mammalian cell that T . gondii infects requires far fewer kinase regulators to complete cell division , which suggests that these parasites have unique vulnerabilities . The increased complexity in parasite cell cycle controls likely evolved from the need to adapt to different hosts and the need to construct the specialized invasion apparatus in order to invade those hosts . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"parasite",
"groups",
"centrosomes",
"parasitic",
"cell",
"cycles",
"toxoplasma",
"gondii",
"cell",
"cycle",
"and",
"cell",
"division",
"cell",
"processes",
"parasitic",
"protozoans",
"parasitology",
"mitosis",
"developmental",
"biology",
"apicomplexa",
"tachyzoites",
"protozoans",
"toxoplasma",
"cellular",
"structures",
"and",
"organelles",
"chromosome",
"biology",
"life",
"cycles",
"parasitic",
"life",
"cycles",
"cell",
"biology",
"biology",
"and",
"life",
"sciences",
"organisms",
"cyclins"
] | 2017 | Checkpoints of apicomplexan cell division identified in Toxoplasma gondii |
Colicins are plasmid-encoded narrow spectrum antibiotics that are synthesized by strains of Escherichia coli and govern intraspecies competition . In a previous report , we demonstrated that the global transcriptional factor IscR , co dependently with the master regulator of the DNA damage response , LexA , delays induction of the pore forming colicin genes after SOS induction . Here we show that IscR is not involved in the regulation of nuclease colicins , but that the AsnC protein is . We report that AsnC , in concert with LexA , is the key controller of the temporal induction of the DNA degrading colicin E8 gene ( cea8 ) , after DNA damage . We demonstrate that a large AsnC nucleosome-like structure , in conjunction with two LexA molecules , prevent cea8 transcription initiation and that AsnC binding activity is directly modulated by L asparagine . We show that L-asparagine is an environmental factor that has a marked impact on cea8 promoter regulation . Our results show that AsnC also modulates the expression of several other DNase and RNase colicin genes but does not substantially affect pore-forming colicin K gene expression . We propose that selection pressure has “chosen” highly conserved regulators to control colicin expression in E . coli strains , enabling similar colicin gene silencing among bacteria upon exchange of colicinogenic plasmids .
Colicins are high-molecular-weight toxic proteins that are produced by and specifically target Escherichia coli and its close relatives [1] . These narrow-spectrum antibiotics kill by either targeting the DNA , RNA or cell membranes of susceptible cells . Cytoplasmic colicins are released upon the synthesis of a lysis protein , the expression of which is independent of intracellular colicin accumulation [2] . This causes the stochastic lysis of producing cells and is suggested to assist surviving sister cells by killing potential competing sensitive cells [3] . Colicin-mediated competition has been suggested to have functions in modulating population dynamics and maintaining diversity of microbial communities [4–7] . Nutrient limitation and DNA damage seem to be the major signals that control colicin production , enabling interference competition among strains [8] . Colicins are plasmid-encoded and are expressed from strong promoters whose activity is tightly repressed by the LexA transcription factor , the master regulator for the SOS DNA damage repair response in bacteria [1 , 9] . Most of the SOS genes involved in DNA repair and cell division arrest are expressed immediately after DNA damage , but induction of colicin genes is delayed . This presumably provides cells time to repair DNA in order to preserve the integrity of their genome , before the induction of colicin production [10] . In previous work , we established that the global transcriptional repressor IscR delays the induction of the pore-forming colicin K gene ( cka ) [11] . We showed that IscR participates in a double-locking mechanism , in concert with LexA , by stabilizing the LexA SOS repressor at the promoter and this links colicin expression to the nutritional status of the cell . Thus , the IscR protein uncouples the induction of colicin expression from the temporal induction of the SOS response that deals with repairable DNA damage . This mechanism also operates at other promoters , which control the expression of bactericidal pore-forming colicins [11] , however , it is not known if a similar fail-safe double-lock system has also evolved for the nuclease colicin genes . Here we report that IscR does not modulate the expression of nuclease colicin genes . Hence , we studied the regulation of the DNA degrading colicin E8 gene ( cea8 ) in more depth and identified the AsnC transcription factor as directly responsible for the delay in cea8 expression . AsnC is a member of Lrp/AsnC family of transcriptional regulators that modulate cellular metabolism in both archaea and bacteria [12 , 13] . In E . coli , AsnC is required to activate the expression of the L-asparagine synthetase A gene ( asnA ) and this stimulation is abolished in the presence of the amino acid L-asparagine [14] . In addition to this , expression of asnC is negatively autoregulated by AsnC and also repressed by the nitrogen assimilation control ( Nac ) protein , under nitrogen-limiting conditions [15] , however , this regulation is not modulated by the presence of L-asparagine [14] . Functional E . coli AsnC is an octamer , whose structure was resolved by X-ray crystallography [13] . We show that AsnC binds to the cea8 regulatory region at multiple sites , likely wrapping the DNA into a nucleoprotein assembly and that its binding is affected by L-asparagine . In the LexA-AsnC-cea8 complex , two LexA dimers are flanked by multiple AsnC octamers , and the presence of L-asparagine influences AsnC modulated promoter region geometry . Data presented here shows that double locking by LexA and AsnC operates at the cea8 promoter region to delay induction of the colicin E8 gene , thereby linking cea8 expression to DNA damage and L-asparagine availability . Thus , AsnC provides colicinogenic cells with time for DNA damage repair and limits colicin E8 induction to terminally damaged cells .
To study the induction of various nuclease colicins after DNA damage , we assayed the activity of colicin promoters in E . coli K-12 strain BW25113 . In this experiment , different DNA fragments , carrying colicin promoters , were cloned into the lac expression vector , pRW50 , to give colicin promoter::lac fusions . After inducing the DNA damage response with a sub-inhibitory concentration of nalidixic acid , we observed that the promoters of the DNA degrading colicins E2 , E7 and E8 , of colicins E5 and D , targeting tRNA , and of the rRNA cleaving colicin E6 , are only induced after a prolonged delay ( Fig 1A ) . Note that there is little expression from nuclease colicin promoters in the absence of DNA damage ( S1 Fig ) . We previously established that the delayed expression of several pore-forming colicins , is due to co-repression by the global transcriptional repressors LexA and IscR [11] . At the colicin K promoter , the LexA repressor was shown to bind to the tandem operators just downstream of the -10 promoter element and prevented RNA polymerase binding . The IscR protein was suggested to increase stability of LexA at these targets . To determine if a similar mechanism controls the expression of the nuclease colicins , we investigated the regulation of the DNA degrading colicin E8 by electrophoretic mobility-shift assays ( EMSA ) and DNase I footprinting . Our results reveal that at the promoter region of cea8 , the LexA repressor binds to two overlapping targets and blocks the access of RNA polymerase to the promoter ( Fig 1B and 1C ) . To investigate the possible binding of IscR to the cea8 promoter , an EMSA assay was again used . Results in Fig 1D show that IscR binds specifically to the cka promoter region but not to that of cea8 . In addition , we tested whether the IscR protein is directly responsible for the delayed production of colicin E8 and other nuclease colicins , by comparing the colicin production in our wild-type and ΔiscR strains . Results , illustrated in Fig 1E , show that IscR has a negligible effect on the synthesis of many nuclease colicins . This contrasts with the pore-forming colicin K , where the ΔiscR allele causes a 100 increase after the first hour of induction ( S3 Fig ) [11] . This indicates that IscR does not regulate any of the nuclease colicin genes and prompted us to search for other transcription factors , involved in controlling the timing of their expression . To investigate the delay in cea8 induction in SOS-induced cells we used a pull-down assay [11] , using a cleared cell extract from mid-logarithmic grown , SOS induced , E . coli cells , and a biotinylated 179 bp cea8 promoter fragment as a bait . Eluted proteins were separated by SDS-polyacrylamide gel electrophoresis and nine bands were analysed by mass spectroscopy ( Fig 2A ) . We identified 30 transcription regulators and nucleoid associated factors that had associated with the bait ( S1 Table ) . To screen for their ability to regulate cea8 expression after DNA damage induction with nalidixic acid , we measured cea8::lac activity in deletion mutants from the Keio collection [16] and we selected strains in which a 3-fold increase in cea8 promoter activity , in comparison to the wild-type strain , was observed ( S2 Table ) . Thus , we focused on the AsnC , StpA , OmpR , YbjK , YihW , YegW and MngR proteins and measured cea8 promoter activity following SOS induction using pRW50 cea8::lac fusion in the corresponding deletion mutant strains throughout the bacterial growth curve . Results presented in Fig 2B show that disruption of asnC resulted in the biggest effects on cea8 promoter induction after DNA damage . An intermediate increase in promoter activity was observed in the strain deficient for stpA , whilst the other deletions had a minimal effect , with our data confirming that IscR does not regulate colicin E8 expression . The StpA protein , a paralogue of the nucleoid-associated protein H-NS , forms a rigid filament along DNA , and can cause DNA bridging [17] . Furthermore , StpA can act as an RNA chaperone [18] and a transcriptional repressor [19 , 20] , thus , it may be involved in colicin gene expression . However , here we focused on AsnC , and assayed its binding to cea8 promoter region and its effect on colicin E8 synthesis . To do this , we introduced the ΔasnC allele into a strain that harbours a cea8-encoding plasmid . After treatment of cells with a subinhibitory concentration of nalidixic acid that induced DNA damage , cell growth and colicin production was compared in the wild-type and the ΔasnC mutant . Our results show that AsnC enhances viability of the strain harbouring the colicin E8-encoding plasmid ( Fig 2C ) . Bioassays were also used to follow colicin levels in crude cell extracts prepared from cells before and after SOS induction . The results show that , in the ΔasnC strain , colicin E8 was produced an hour earlier in comparison to the delayed synthesis in the wild-type strain ( Fig 2D ) . This suggests that AsnC directly represses cea8 promoter activity and , in concert with LexA , ensures regulated and delayed expression of the cea8 gene . The AsnC protein is a member of the Lrp/AsnC family of regulators , which often assemble to form wheel-like octamers and whose DNA binding activity can be modified by small molecules , such as amino acids [13] . AsnC regulates the expression of its own gene , asnC , and the asnA gene , encoding for a synthetase that catalyses the ammonia-dependent conversion of aspartate to asparagine [14] . To investigate the binding of AsnC to the cea8 promoter ( Fig 3A ) , we over-expressed and purified the AsnC protein and performed in vitro experiments in the presence or absence of the amino acid L-asparagine . EMSA experiments show that several AsnC molecules can interact with cea8 ( Fig 3 ) and that in the presence of L-asparagine a number of distinct complexes can be observed ( Fig 3B ) . In the absence of L-asparagine , at higher AsnC concentrations , DNA remained in the wells of the gel , indicating that high molecular weight nucleoprotein complexes had formed . DNase I footprinting was also used to study the location of AsnC binding to the cea8 promoter sequence , again in the presence or absence of L-asparagine . Results in Fig 3C show that AsnC interacts along the entire length of the 179 bp cea8 promoter region . Inspection of the cea8 region , interacting with AsnC , revealed 27 DNase I hypersensitive sites , which is indicative of local bending and distortion of the DNA helix . This results in a widening of the minor groove and makes the DNA more susceptible to DNase I attack , leading to the production of hypersensitive bands [21] . In several locations the presence of L- asparagine altered the binding of AsnC to the cea8 promoter ( see red boxes in Fig 3C ) . The position of the red boxes in Fig 3C was determined by comparing the AsnC footprint gels in the presence or absence of L-asparagine . Our in vitro analysis indicates that AsnC binds to the cea8 promoter region at multiple sites , likely wrapping the DNA into a complex nucleoprotein assembly , and that the architecture of this complex is altered by the presence of L-asparagine ( Fig 3 ) . Since our data show that AsnC binds at multiple locations and alters the architecture of colicin E8 regulatory region , as well as binding within LexA target sites ( Fig 3 ) , we tested if AsnC and LexA can simultaneously bind to the cea8 promoter region . To investigate this , we performed EMSA analysis on the cea8 promoter fragment . Using purified LexA and AsnC , in the presence of L-asparagine , we observed a large nucleoprotein complex composed of at least two AsnC functional oligomers , presumably octamers [13] , and two LexA dimers interacting at cea8 ( Fig 4A ) . Note , that LexA was used at a concentration of 400 nM at which LexA repressor occupies both LexA binding sites within cea8 ( Fig 1B ) . To determine whether occupancy of the DNA by AsnC affects the binding of LexA at the cea8 promoter region , we performed DNase I footprint analysis and compared signatures of LexA and AsnC in the presence or absence of L-asparagine . In both conditions , LexA repressors bound to tandem targets just downstream of the -10 promoter element ( Fig 4B ) . As observed for AsnC binding in Fig 3C , the addition of L-asparagine also modulated the binding of AsnC in the LexA-AsnC nucleoprotein complex ( Fig 4B and 4C ) . In the AsnC-LexA-cea8 complex , specific hypersensitive sites were apparent ( determined by stars in Fig 4B ) , suggesting that the binding of both proteins subtly alters the structure or trajectory of the DNA around -10 element ( Fig 4D–4F ) . Thus , we conclude that concurrent binding of LexA and AsnC to the cea8 regulatory region ensures delayed induction of the DNase E8 synthesis after DNA damage . Our data suggest that tight repression of DNase colicin E8 might be affected by the availability of amino acid L-asparagine and this signal is relayed via AsnC . To test this hypothesis we measured cea8 promoter activity in SOS-induced wild-type cells grown in the M9 minimal medium containing either 10 mM NH4Cl or 20 mM L-asparagine as the sole source of nitrogen . Fig 5 shows that in the L-asparagine containing medium , the expression from cea8 remains low , whilst it increases in medium containing higher levels of NH4Cl . This is in agreement with our in vitro data ( Figs 3 and 4 ) and suggests that L-asparagine is needed to stabilize a specific AsnC assembly at the cea8 promoter region . Hence , we suggest that depletion of L-asparagine is the signal for AsnC de-repression at the cea8 promoter . To determine whether AsnC modulates the expression of other colicins , we assayed DNase colicin E2 ( cea2 ) and rRNase colicin E6 ( cea6 ) promoter activities following SOS induction with nalidixic acid using cea6::lacZ and cea2::lacZ promoter fusions in the wild-type , ΔiscR and ΔasnC strains . Results illustrated in Fig 6A and 6B show that the disruption of asnC resulted in elevated cea2 and cea6 promoter activity immediately after DNA damage induction , when compared to the wild-type and the ΔiscR strains . This indicates that AsnC , rather than IscR , is a key transcriptional repressor of the cea2 and cea6 promoters . In addition , we transferred the colicinogenic plasmids for these DNase and RNase colicins into the ΔasnC and wild-type strains . Following the induction of DNA damage , cell growth ( Fig 6C ) and colicin production ( Fig 6D ) was monitored in both strains . In the absence of asnC , cells failed to reach as high optical density as in the wild-type strain , suggesting that elevated colicin expression and cell lysis had taken place ( Fig 6C ) . Cell extracts were prepared from cultures before and after DNA damage induction and colicin levels compared by a colicin production bioassay ( Fig 6D ) . After SOS induction , nuclease colicin synthesis was induced earlier for colicins E2 , E5 and E6 in the ΔasnC strain , in comparison to the wild-type strain , indicating that AsnC directly modulates the expression of a number of other colicin genes . Alignment of the cea8 promoter region sequence with corresponding sequences from colicin E2 , E5 and E6 indicated that these promoters are very similar ( S4 Fig ) and , thus , similar co-ordinated regulation is perhaps to be expected . It is clear from S4 Fig that regions of the cka promoter are similar to that of cea8 , particularly around the two LexA SOS boxes . As this region was bound differentially by AsnC at cea8 in the presence of L-asparagine , we examined whether purified AsnC could bind in vitro to a radiolabelled cka promoter fragment using EMSA . Results in S5A Fig indicated that AsnC does bind to cka and its DNA binding was modulated by L-asparagine . As this raises the possibility that AsnC could regulate colicin K production in vivo , we measured cka promoter activity after SOS-induction in wild-type , ΔasnC and ΔiscR cells , carrying a cka::lacZ fusion cloned into pRW50 . Results in Fig 6E indicate that AsnC had little effect on cka expression , but confirmed that IscR is a major repressor of the cka promoter . In addition , colicin K expression was also examined in our wild-type , ΔasnC and ΔiscR strains , whilst carrying a colicin K-encoding plasmid ( S5B and S5C Fig ) . These experiments again showed that IscR is the major regulator of colicin K expression and that AsnC has little effect on the expression of this pore-forming colicin . E . coli harbours many promoters that are regulated by multiple transcription factors , each of which ensures that different intra- or extracellular signals are integrated into gene expression [22] . At the DNase colicin E8 regulatory region we identified a large nucleoprotein complex composed of two LexA repressors flanked by at least two AsnC octamers , that likely wrap DNA in a nucleosome-like structure to firmly prevent transcription initiation ( Fig 4D–4F ) . The AsnC protein belongs to the Lrp/AsnC family of transcriptional regulators that are widely distributed among prokaryotes and affect cellular metabolism , often in response to exogenous amino acids [12 , 13] . In contrast to other members of the family , which are global regulators and affect a variety of bacterial functions [23] , the AsnC protein was thought to be a gene specific regulator , controlling only two genes in E . coli ( asnA and asnC ) [14] . Here we report a novel role for AsnC , in which the promoters of the nuclease colicins have “recruited” this protein , enabling regulation in response to L-asparagine levels . Our data show that an amino acid effector modulates AsnC interaction at the colicin E8 promoter , which influences regulation of cea8 expression . We predict that , as for the Neisseria meningitidis AsnC ortholog , [24] , L-asparagine binding modulates the stability of a certain protein oligomeric state and also the mode of binding in the E . coli AsnC-cea8-LexA nucleoprotein complex . Furthermore , as expression of asnC is negatively autoregulated and dependent on the Nac protein under nitrogen-limiting conditions [15] , nutrient conditions , specifically nitrogen levels , and nitrogen metabolism might coordinate the cea8 expression through altering the amount of AsnC within the cell . Note that nutrients were recently reported to modulate the release of the DNase colicins by modulating the translation efficiency of the colicin E2 lysis gene transcript [25] . Therefore , AsnC appears to couple metabolic signals to the induction of colicin operon components , in order to synchronise accumulation with the release of the colicin . In our previous work , we showed that the global transcriptional factor , IscR , in response to the nutritional status of the cell , and , co-dependently with LexA , delays induction of the pore-forming colicin genes after SOS induction [11] . This was a surprising finding as E . coli IscR had been thought to be primarily involved with controlling housekeeping iron sulphur cluster biogenesis , anaerobic respiration enzymes and biofilm formation [26 , 27] . Here our data strongly suggests that temporal induction of DNA and RNA targeting colicins is IscR independent , and show that the key regulator is the AsnC repressor . At the cea8 promoter , AsnC repression seems to reflect L-asparagine levels and presumably serves as an indicator of general amino acid abundance and availability . In contrast , AsnC does not affect the expression of pore-forming colicin K gene expression . Thus , our data imply that the promoters of the nuclease and pore-forming colicins have adopted different transcription regulators to co-ordinately regulate transcription in conjunction with the LexA repressor and distinct metabolic inputs are integrated at these promoters , which both affect the timing and level of colicin induction . It is clear that colicinogenic plasmids seem to have evolved to exploit transcriptional factors that are of the host origin . We suggest that ubiquitous regulators , present in most E . coli strains were “picked” in order that the colicinogenic plasmids can be swapped between strains [1] , with the colicin promoters being silenced in the same manner . In addition , colicin production and subsequent lysis protein driven colicin release causing death of the producing bacteria , may enable eradication of strains that lack or synthesize a non-functional regulator and cannot efficiently respond or adapt to different environmental signals .
The bacterial strains , plasmids , promoter fragments and oligodeoxynucleotide primers used in the present study are listed in S3 Table . The E . coli Keio collection wild-type strain , BW25113 , and its derivatives were used throughout the study [16] . To verify the Keio collection deletion strains , a transposon-specific primer Keio1Kn [16] and the gene specific primer ( named as gene pre ) was used ( S3 Table ) in 30 cycles of PCR reactions ( 30 sec 94°C , 30 sec 55°C , 60 sec 72°C ) . PCR products were analysed on 1 . 2% agarose gels and stained with ethidium bromide . The colicin D ( cda ) , E2 ( cea2 ) , E5 ( cea5 ) , E6 ( cea6 ) , E7 ( cea7 ) and E8 ( cea8 ) promoter fragments were amplified by PCR from natural colicin encoding plasmids using primers colX_beta_F and colX_beta_R ( X denotes the relevant colicin ) , which introduce flanking EcoRI and HindIII sites ( S3 Table ) . For testing colicin promoter activities , each promoter fragment was cloned into the lac expression vector , pRW50 . Plasmid constructs were named as pRW50cxay ( x and y denotes each colicin ) . As a source of DNA fragments for in vitro analysis , EcoRI-HindIII colicin E8 and colicin K promoter fragments were cloned into pSR . To assay colicin synthesis and ensure plasmid selection , the transposon Tn3 ( ApR ) was inserted into the naturally occurring colicinogenic plasmids of the Pugsley colicin collection [28] , harbouring operons for either colicin D , E2 , E5 , E6 , E7 or E8 . Strain CL127 carrying Tn3 on the conjugative plasmid pHly152-T8 was used as a donor strain . To generate plasmid pAsnC , for the overexpression of the N-terminal His6 AsnC fusion protein , primers asnC_u and asnC_d were used to PCR amplify the asnC open reading frame and introduce flanking BamHI and MluI restriction sites . Purified PCR product was subsequently cloned into expression vector pET8c ( Novagen ) to generate pAsnC . E . coli RNA polymerase holoenzyme harbouring σ70 ( RNAP ) was obtained from Epicentre Technologies ( Madison ) . The His6-LexA protein was overexpressed and purified as described in [29] and stored in 20 mM Tris ( pH 7 . 3 ) , 200 mM NaCl at -80°C . The His6-IscR protein was overexpressed , purified and its concentration determined as described in [11] . To induce the synthesis of AsnC protein , an overnight culture of E . coli BL21 ( DE3 ) pLysE strain grown on an agar plate , containing ampicilin ( 100 μg ml-1 ) and chloramphenicol ( 25 μg ml-1 ) , harbouring pAsnC was grown to an optical density at 600 nm ( OD600 ) of 0 . 6 when 0 . 8 mM isopropyl β-D-1-thiogalactopyranoside ( IPTG ) was added to the culture . After 4 h of growth the cells were harvested and the N-terminally His6-tagged AsnC was affinity purified by Ni-chelate chromatography ( Quiagen ) and stored at 4°C in 50 mM NaH2PO4 ( pH 8 ) , 300 mM NaCl , 250 mM imidazole . The concentrations of the LexA and AsnC proteins were determined using a NanoDrop 1000 ( Thermo Scientific ) using the extinction coefficients at 280 nm of 6990 M-1 cm-1 and of 10555 M-1 cm-1 , respectively . The low-copy number lac expression vector , pRW50 [30] , was used to measure the activity of the colicin promoters . Plasmids harbouring colicin promoter fragments ( S3 Table ) were transformed into the relevant strains . Cells were grown aerobically ( 180 r . p . m . ) at 37°C in Lysogeny Broth ( LB ) supplemented with tetracycline ( 12 . 5 μg ml-1 ) . To induce the SOS response , a sub-inhibitory concentration [31] , 37 μM , of nalidixic acid ( Sigma-Aldrich ) was added to the culture when the OD600 reached 0 . 3 . Culture samples were assayed for β-galactosidase activity according to the Miller method [32] . The presented values are the averages of at least three independent experiments and are shown with standard deviations . To measure the L-asparagine effect on cea8 promoter activity , the relevant wild type strain carrying pRW50cea8 was grown to an OD600 ~0 . 2 in M9 medium [33] , containing a low concentration of NH4Cl ( 0 . 5 mM ) to sustain growth . The bacterial culture was then split and supplemented with either 10 mM NH4Cl or 20 mM L-asparagine . After the addition of 37 μM nalidixic acid , as indicated , samples were taken and analysed as described above . A biotinylated 179 bp colicin E8 promoter fragment from position -169 to position +10 from the translation start site ( TSS ) was generated by PCR using primers Pull_FE8 and Pull_RE8 with pColE8-Tn3 as a template . The DNA fragment was purified by GeneJET PCR purification kit ( Thermo Scientific ) . Immobilisation of the biotinylated DNA ( 50 μg ) to 5 mg of M-280 streptavidin Dynabeads ( Invitrogen ) was carried out in 15 minutes at room temperature as described [11] . An overnight culture of the E . coli BW25113 , harbouring the pRW50cea8 , was diluted 1: 200 into 0 . 5 l LB broth supplemented with tetracycline ( 12 . 5 μg ml-1 ) and induced with nalidixic acid ( 37 μM ) once the OD600 had reached 0 . 3 . After 45 min , cells were harvested and cell extract prepared as described [31] . Cleared lysates ( ~20 ml ) were mixed with streptavidin beads with or without cross-linked biotinylated cea8 promoter fragment in 50 ml centrifuge tubes ( Costar ) and incubated for 10 min with gentle mixing on ice . Dynabeads were collected using a magnet and washed four times in 20 mM Hepes-Na ( pH 7 . 4 ) , 100 mM NaCl , 0 . 1% ( v/v ) Tween 20 . Proteins were eluted with 500 μl of buffer ( 20 mM Hepes-Na , 800 mM NaCl , 0 . 1% ( v/v ) Tween 20 ) and concentrated by TCA precipitation . Proteins were resolved on a 12% SDS-PAGE gel ( Invitrogen ) , and visualized by Coomassie blue staining . To identify proteins , nine 1 mm gel slices were excised and analysed by the Functional Genomics , Proteomics and Metabolomics Facility at the University of Birmingham using a Thermo-Finnigan LTW Orbitrap mass spectrometer . Candidate proteins that exhibited DNA binding properties were analysed further . Colicin synthesis was monitored in the wild-type E . coli BW25113 strain or its ΔasnC derivative JW3721 [16] , harbouring one of the colicinogenic plasmids , and was grown aerobically at 37°C in LB broth supplemented with ampicillin . Nalidixic acid ( Sigma-Aldrich ) was added to the culture at a final concentration of 37 μM , when the OD600 reached 0 . 3 . Samples were taken before induction and at 1 , 2 and 3 h after . Cells were diluted to obtain 1 ml samples with an OD600 of 0 . 3 and the crude cell extracts were prepared by sonication and the cell debris cleared by centrifugation for 1 min at 17000 x g . 100 μl of each extract was injected into wells in an agar plate containing tetracycline ( 12 . 5 μg ml-1 ) overlaid with the lawn of the indicator strain ( DH5α harbouring pBR322 ) as described in [11] . As an alternative approach for colicin determination in the crude cell extracts , 5 μl of a ten-fold or five-fold dilution series of extracts were applied to an agar plate overlaid with the indicator strain as above . Indicator strains were grown at 37°C and the plates photographed using a G:Box ( Syngene ) . EMSA analysis , using purified LexA , IscR , AsnC and RNAP , with the cea8 and cka promoter regions , was performed as described in [34] . DNA fragments were excised from pSRcea8 or pSRcka using EcoRI and HindIII restriction enzymes and purified promotror fragments were labelled at the HindIII end with [γ-32P]-ATP using polynucleotide kinase ( NEB ) . Approximately 0 . 5 ng of DNA fragment was incubated with varying amounts of purified proteins , as indicated . The reaction buffer contained 20 mM Hepes ( pH 8 ) , 5 mM MgCl2 , 50 mM potassium glutamate , 1 mM DTT , 5% ( v/v ) glycerol and 0 . 5 mg/ml BSA and the final reaction volume was 10 μl . Where AsnC was used , the EMSA buffer contained 5 mM L-asparagine ( L-asn ) , where indicated . Samples were incubated at 37°C for 15 min before electrophoresis . For competitive EMSA experiments , DNA fragments were first incubated with various concentrations of LexA ( for 15 min at 37°C ) followed by the addition of RNAP and incubated for another 15 min at 37°C . Herring sperm DNA was included at a concentration of 6 . 5 μg ml-1 for these experiments . After incubation , all samples were immediately run on a 5% polyacrylamide gel at 12 V cm-1 in 0 . 25 x TBE , running under tension , and were visualised using a Bio-Rad Molecular Imager FX and Quantity One Software ( Bio-Rad ) . DNase I footprinting of AsnC and LexA at the cea8 promoter region was performed as described [35] , using purified proteins in the presence or absence of 5 mM L-asparagine and a purified EcoRI-HindIII cea8 fragment that had been 32P-end labelled at the HindIII site using polynucleotide kinase and [γ-32P]ATP . | Colicins are considered model proteins for studying bacterial toxins . These narrow spectrum antibiotics can kill by a variety of mechanisms , e . g . by forming pores in the membranes of susceptible cells or by degrading their nucleic acids . Colicin genes are plasmid-encoded and repressed by the master regulator of the DNA damage response , LexA . Induction of several pore-forming colicin genes is also repressed by IscR , which ensures that colicin genes are switched on as a last resort in DNA damaged cells , when nutrients are depleted . Here we show that nuclease colicin genes are not controlled by IscR but that the AsnC protein , in concert with LexA , is directly responsible for uncoupling the immediate expression of the DNase colicin E8 from the main induction of the SOS response . AsnC wraps the DNA of the colicin E8 promoter into a complex nucleoprotein assembly and the architecture of this complex is altered by the presence of the amino acid L-asparagine . Thus , repression by metabolite-responsive and DNA-damage responsive regulators operates at the regulatory regions of different colicins . Hence , the response to several environmental signals have been integrated to ensure that , following DNA damage , colicin synthesis is tightly repressed and induced only in terminally damaged cells . | [
"Abstract",
"Introduction",
"Results",
"and",
"Discussion",
"Material",
"and",
"Methods"
] | [] | 2015 | Silencing of DNase Colicin E8 Gene Expression by a Complex Nucleoprotein Assembly Ensures Timely Colicin Induction |
Intestinal stem cell ( ISC ) self-renewal and proliferation are directed by Wnt/β-catenin signaling in mammals , whereas aberrant Wnt pathway activation in ISCs triggers the development of human colorectal carcinoma . Herein , we have utilized the Drosophila midgut , a powerful model for ISC regulation , to elucidate the mechanisms by which Wingless ( Wg ) /Wnt regulates intestinal homeostasis and development . We provide evidence that the Wg signaling pathway , activation of which peaks at each of the major compartment boundaries of the adult intestine , has essential functions . Wg pathway activation in the intestinal epithelium is required not only to specify cell fate near compartment boundaries during development , but also to control ISC proliferation within compartments during homeostasis . Further , in contrast with the previous focus on Wg pathway activation within ISCs , we demonstrate that the primary mechanism by which Wg signaling regulates ISC proliferation during homeostasis is non-autonomous . Activation of the Wg pathway in absorptive enterocytes is required to suppress JAK-STAT signaling in neighboring ISCs , and thereby their proliferation . We conclude that Wg signaling gradients have essential roles during homeostasis and development of the adult intestine , non-autonomously controlling stem cell proliferation inside compartments , and autonomously specifying cell fate near compartment boundaries .
The evolutionarily conserved Wnt/β-catenin signal transduction pathway regulates cell proliferation and tissue patterning in metazoans , and deregulation of this pathway is associated with numerous human diseases [1 , 2] . In the absence of Wnt/Wg exposure , the key transcriptional activator β-catenin/Armadillo ( Arm ) is targeted for proteasomal degradation by a “destruction complex” comprised of Axin , Adenomatous polyposis coli ( Apc1 and Apc2 ) and glycogen synthase kinase 3 ( GSK3 ) /shaggy ( sgg ) . Binding of Wnt/Wg to the transmembrane co-receptors Frizzled ( Fz and Dfz2 ) and low-density lipoprotein receptor-related protein 5/6 ( LRP6 ) /Arrow ( Arr ) recruits the adaptor protein Dishevelled ( Dvl/Dsh ) and deactivates the destruction complex . Stabilized β-catenin subsequently translocates to the nucleus , and associates with the transcription factor T-cell factor/lymphoid enhancer factor ( TCF/LEF ) and the cofactors Pygopus ( Pygo ) and BCL9/Legless ( Lgs ) to activate target genes ( S1A Fig ) [3–7] . Notably , Wnt signaling is essential for self-renewal and proliferation of mammalian ISCs , whereas Wnt pathway hyperactivation triggers the development of the vast majority of colorectal carcinomas [2 , 8] . The Drosophila digestive tract , with its remarkable similarity to the mammalian intestine but simpler anatomy , is an ideal model for studying intestinal development , homeostasis , and disease [9–12] . Like its mammalian counterpart , the fly gut undergoes rapid turnover and is replenished by ISCs . The ISCs , which are distributed along the basement membrane of the gut epithelium , divide asymmetrically to give rise to enteroblasts and pre-enteroendocrine cells that differentiate into either absorptive enterocytes or secretory enteroendocrine cells , respectively [13–17] . Visceral muscles envelop this monolayer epithelium . Food is successively digested , absorbed and eliminated through the foregut , midgut and hindgut . The midgut can be further partitioned into fine-scale compartments of unique histological structure , gene expression , and physiological function [18–20] , denoted as R1 ( Region 1 ) through R5 ( Fig 1A ) . Intriguingly , the peak expression of frizzled 3 ( fz3 ) , a direct target gene of the Wg pathway [21 , 22] , occurs at four of these compartment boundaries ( cardia-R1 , R2-R3 , R3-R4 and R5-hindgut ) [18] . Wg signaling is required for intestinal regeneration after injury caused by toxins or bacterial infection , whereas the aberrant activation of signaling deregulates ISC proliferation [23–29] . The roles of Wg signaling during homeostasis have also been examined previously . An initial study indicated that wg expression in the muscle surrounding the midgut activates the pathway in midgut ISCs , and is required for their self-renewal and proliferation [30] . However , later studies challenged these conclusions [25 , 26] . First , ISC self-renewal was not affected upon loss of Apc [26] . Second , knockdown of wg from both muscle and epithelial sources , or in wgCX4 heterozygous mutants did not lead to significant loss of ISCs even after 30 days [25] . Third , genetic inactivation of core Wnt pathway components with null alleles resulted in only mild or no effects on ISC proliferation during homeostasis [25] . The only method that revealed strong effects of Wg signaling on ISC self-renewal [25 , 26 , 30] required ectopic expression of a dominant negative dTCF protein [4] . The recent discovery that Wg signaling is enriched specifically at compartment boundaries , as revealed by activation of the target gene fz3 [18] , prompted us to reexamine the source and roles of Wg during adult intestinal homeostasis and development . Here , we identify several novel sources of Wg in the intestinal epithelium and in the surrounding visceral muscle , which all peak at compartment boundaries . We confirm that Wg pathway activation also peaks at these boundaries , but also find that low-level Wg signaling is present throughout compartments , where it is essential for maintenance of homeostasis . Further , in contrast with the prior focus on Wg signaling in midgut stem cells , our findings reveal that enterocytes , and not ISCs , are the primary site of Wg pathway activation during homeostasis . Wg signaling in enterocytes non-autonomously regulates JAK-STAT signaling in neighboring ISCs , and thereby prevents ISC overproliferation during homeostasis . Further , we demonstrate that Wg signaling is required for proper cell fate specification near compartment boundaries during development . We conclude that gradients of Wg signaling that peak at adult intestinal compartment boundaries are essential to control stem cell proliferation during homeostasis and to specify cell fate during development .
To identify regions in the adult intestine that express wg , we examined a wg-lacZ enhancer trap [31] ( Fig 1B ) . Consistent with prior findings , four rows of wg-expressing cells were detected in the surrounding circular visceral muscles throughout the entire length of the midgut [30] ( Figs 1D and S1C ) . Furthermore , at the foregut-midgut boundary , wg was expressed in the anterior cardiac epithelium , whereas at the midgut-hindgut boundary , wg was expressed in the spindle cells of the hindgut proliferation zone ( HPZ ) [23 , 32–35] ( Figs 1C , 1E and S1B ) . Unexpectedly , we also identified four novel sources of Wg at intestinal compartment boundaries ( Figs 1F–1I” and S1D–S1D” ) . First , we observed high-level wg expression in a band of approximately 18 visceral muscle cells in the middle of the midgut , adjacent to the anterior border of the copper cell region ( CCR , marked by Cut and α-Spectrin [24] ) and the R2-R3 compartment boundary ( Fig 1B and 1F-1G” ) . Second , we identified a zone of wg expression in the visceral muscle at the posterior terminal midgut , immediately anterior to the R5-HPZ border [23 , 34] ( Fig 1H–1H” ) . Third , we detected high-level wg expression in the intestinal epithelium underneath this wg-enriched muscle segment , which was present not only at the R5-HPZ boundary , but also extended anteriorly into the posterior midgut ( Fig 1I–1I” ) . Lastly , a transverse stripe of wg was present within the rectal epithelium ( S1D-S1D” Fig ) [36] . Together , these results indicated that wg is not expressed uniformly along the fly gut , but is enriched at intestinal compartment boundaries . Notably , this expression pattern was largely retained in aged flies , indicating that the enrichment of wg expression at major compartment boundaries was stable over the course of adulthood ( S1E–S1E” Fig ) . These unprecedented observations of boundary-enriched wg expression were recapitulated using a wg:mcherry knock-in line [37] ( S2 Fig ) , and when lacZ expression was driven by a wg:Gal4 knock-in line [37] ( S3 Fig ) . Of note , expression of wg>lacZ culminated at all major compartment boundaries ( S3 Fig ) and the epithelial sources were detected in enterocytes ( S4 Fig ) . In summary , the newly discovered sources in the epithelium and muscle , coupled with previous findings , revealed marked enrichment of wg expression at adult intestinal compartment boundaries that correlate with major sphincters and tissue-organizing centers . Having established the sources of Wg that are present in the epithelium and visceral muscle of the adult midgut , we sought to identify the regions where the Wg pathway is activated during homeostasis . We confirmed that a direct target gene of Wg signaling , frizzled 3 ( fz3 ) [21 , 22 , 38] , was expressed in five gradients in the adult gut , each of which peaked at intestinal compartment boundaries , as described previously [18] ( Fig 2A–2D and 2E–2F ) . We also detected strong fz3-RFP signal at the anterior border of the rectal papillae ( S5A and S5A” Fig ) . To validate these conclusions , we analyzed a reporter for another direct Wg pathway target gene , naked ( nkd ) [39] . As observed for fz3-RFP , nkd-lacZ was also expressed highly at intestinal compartment boundaries , though in a narrower pattern ( Fig 2A and 2B’–2F ) . Further , nkd-lacZ was also expressed in the rectum ( S5A’ and S5A” Fig ) , and , as observed for wg , in four rows in the visceral muscle surrounding the intestine ( Fig 2A ) . Notably , we also detected expression of both fz3-RFP and nkd-lacZ at low levels within posterior midgut compartments ( Fig 2A and 2E–2F ) . To determine whether this low-level expression was dependent on Wg signaling , we inactivated the pathway by generating MARCM clones [40 , 41] of null alleles of the Wg pathway components pygo and dsh [6 , 42] alongside wild-type control clones . The expression of fz3-RFP and nkd-lacZ were eliminated in mutant clones , but not in wild-type clones , indicating that the Wg pathway is indeed activated at low levels in midgut compartments ( Fig 2G–2J’ ) . Together , these findings indicated that the Wg pathway is activated at high levels at the boundaries between functionally distinct intestinal compartments , but also activated at low levels within these compartments ( Fig 2K ) . Recent studies have indicated that the larval intestine is also compartmentalized [18 , 19 , 43] . We therefore sought to determine whether the boundary enrichment of wg and Wg pathway activity are also present in the larval counterpart , and thus examined the expression pattern of wg-lacZ , wg>lacZ and fz3-RFP in 3rd instar larval guts . Remarkably , as in adults , wg expression that emanates from muscle and epithelial sources ( S6 Fig ) was enriched at compartment boundaries of the larval intestine . Further , fz3-RFP expression also culminated at these boundaries ( S7 Fig ) . These findings suggested that even though the larval and adult intestines are derived independently and face different functional demands , the patterns of boundary-enriched Wg pathway activation are largely shared . To further determine whether Wg signaling was required for fz3-RFP expression in the larval gut , we induced either null mutant clones of Wg pathway components , or wild-type control clones , during larval development . Inactivation of Wg signaling resulted in specific loss of fz3-RFP within AMPs at compartment boundaries ( S8A–S8D”‘ Fig ) . Moreover , inactivation of the Apc-Axin destruction complex resulted in the ectopic expression of fz3-RFP in AMPs but not enterocytes [44] ( S8E–S8E”‘ Fig ) . We conclude that in the larval midgut , the Wg pathway is activated exclusively in progenitor cells , and this activation is restricted to compartment boundaries . To test whether stem cells are also the primary sites of Wg pathway activation in the adult intestinal epithelium , we first identified the cell types in which the two Wg pathway reporters , fz3-RFP and nkd-lacZ , were expressed ( Fig 3A–3B”‘ ) . We found that nkd-lacZ was expressed primarily in enterocytes ( demarcated by their large polyploid nuclei ) and also in a subpopulation of enteroendocrine cells ( small diploid cells that are Escargot negative and Prospero positive [13] ) ( Fig 3A–3A”‘ ) , whereas fz3-RFP was detected not only in enterocytes , but also in progenitor cells ( small diploid Escargot positive cells [13] ) ( Fig 3B–3B”‘ ) . Thus , unexpectedly , despite the overlapping boundary-enriched pattern , the expression of the two Wg pathway reporters was partially distinct with regard to individual cell types . We therefore sought to determine whether expression of the reporters was dependent on Wg signaling by generating marked null mutant clones of four Wg pathway components: Dsh , Pygo , Arr [45] and the functionally redundant Fz and Dfz2 [46] , alongside wild-type controls . Nkd-lacZ and fz3-RFP expression were not affected in wild-type control clones ( Figs 2G–2H’ , 3C–3C”‘ and 3E–3E”‘ ) . In contrast , both inside compartments and at compartment boundaries , the inactivation of Wg signaling in the mutant clones eliminated expression of both nkd-lacZ ( Figs 2J–2J’ , 3D–3D”‘ , S9A–S9A”‘ and S10A–S10A”‘ ) and fz3-RFP ( Figs 2I–2I’ , 3F–3G”‘ , S9B–S9E”‘ and S10B–S10C”‘ ) in enterocytes , revealing dependence on Wg pathway activation in this cell type . Unexpectedly , in mutant clones , there was no decrease in the level of nkd-lacZ expression in the subpopulation of enteroendocrine cells ( Figs 3D–3D”‘ and S10A–S10A”‘ ) , indicating that the nkd-lacZ reporter expression in enteroendocrine cells was independent of Wg pathway activation . Further , there was no decrease in the level of fz3-RFP expression in the vast majority of progenitor cells ( Figs 3F–3F”‘ , S9B–S9D”‘ and S10B–S10C”‘ ) , suggesting that the fz3-RFP reporter expression in nearly all progenitors was independent of Wg pathway activation . The only exception was inside R5 , where both enterocytes and progenitors were responsive to Wg exposure , as indicated by loss of fz3-RFP expression in the mutant clones of Wg pathway components in this subregion ( Figs 3G–3G”‘ and S9E-S9E”‘ ) . We conclude that , in contrast with the larval gut , the primary site of Wg pathway activation during adult homeostasis is in enterocytes , and not ISCs , throughout the majority of the midgut . The pattern of wg expression cannot account for the preferential activation of signaling specifically in enterocytes . Therefore , we sought to determine whether intrinsic differences in the distinct intestinal cell types could explain their differential response to Wg exposure . We examined the response of the different cell types to ectopic pathway activation by analyzing the expression patterns of fz3-RFP in Apc1 null mutant flies or in Apc2 Apc1 double mutant clones [47] , in which the Wg pathway is constitutively activated due to loss of destruction complex activity . Compared with wild-type , fz3-RFP expression was markedly increased in Apc1 mutant guts and in Apc2 Apc1 double mutant clones ( Figs 4A–4C”‘ and S11 ) . This aberrant Wg pathway activation was not restricted to enterocytes , but instead observed in all cell types in the intestinal epithelium . These findings indicated that the Wg pathway components acting downstream of the Apc-Axin destruction complex are present in all cell types in the intestinal epithelium , and thus do not underlie their differential responsiveness to Wg stimulation . To determine whether Wg pathway components that act upstream of the destruction complex are functional in all cell types of the intestinal epithelium , we expressed wg throughout the muscle using a temperature sensitive dMef2-Gal4 driver [48] . As compared with controls ( Fig 4D ) , overexpressing wg in the muscle markedly increased the fz3-RFP signal throughout the entire length of the intestine , and most pronouncedly in the anterior midgut ( Fig 4D’ ) . These results indicated that Wg originating from the visceral muscle is sufficient to activate signaling in the intestinal epithelium . Notably , this ectopic activation of Wg signaling was observed in all intestinal cell types ( Fig 4E–4E”‘ ) . Thus , all cell types in the intestinal epithelium express the pathway components necessary for transduction of the Wg signal , and are thus capable of responding to Wg exposure . Therefore , the activation of signaling primarily in enterocytes during homeostasis may reflect inherent differences in the threshold for pathway activation ( see discussion ) . Having discovered that enterocytes are the primary sites of Wg pathway activation in the homeostatic adult gut , we sought to determine whether this signaling was required to maintain homeostasis . We inactivated the Wg pathway by inducing arr , pygo or dsh null mutant clones during early adulthood and examined the effects 5 to 7 days later . We found that wild-type control clones were comprised primarily of 1 to 2 cells ( Fig 5A ) , and were surrounded by regularly spaced progenitor cells . In contrast , a higher percentage of multi-cellular clones of Wg pathway mutants were detected ( Fig 5A ) . Furthermore , an increased number of wild-type progenitors ( esg-lacZ marked cells or small Armstrong Prosperoneg cells ) were present in clusters adjacent to the mutant clones , whereas progenitor cells located farther away from the clones exhibited normal spacing and number ( Figs 5B and S12A–S12B” ) . Further , Dl-lacZ and Su ( H ) -lacZ expression revealed that these clusters contained an increased number of both ISCs and EBs ( Fig 5C–5F ) . We further sought to determine whether the aberrantly increased number of progenitors adjacent to Wg pathway mutant clones resulted from their overproliferation . To test this , we compared the mitotic index in posterior midguts bearing either wild-type or Wg pathway mutant clones . Indeed , many more phospho-histone H3 ( pH3 ) positive cells were observed in posterior midguts containing pygo or dsh clones by comparison with controls ( Figs 5G and S12C and S12D ) . Thus , non-autonomous ISC overproliferation was detected . There remained the possibility that the normal process of ISC migration following mitosis was also disrupted , and contributed to their aberrant clustering . This non-autonomous overproliferation defect was observed only when the Wg pathway was inactivated during adulthood , but not prior to eclosion ( S10 Fig ) . Together , these findings indicate that Wg signaling prevents the non-autonomous overproliferation of neighboring ISCs during adult homeostasis . To test the possibility that Wg signaling in enterocytes non-autonomously regulates the proliferation of neighboring progenitor cells , we disrupted signaling by knocking down Wg pathway components using RNA-mediated interference ( RNAi ) , or by expressing dominant-negative Legless ( Lgs17E ) [5] or dominant-negative TCF ( dTCFΔN ) in enterocytes or progenitor cells with the cell type-specific MyoIA or esg drivers , respectively [13 , 49 , 50] . Inhibition of Wg signaling in enterocytes resulted in ISC overproliferation , as revealed by the aberrantly increased number of pH3+ cells ( Fig 5H ) and the presence of progenitor cells that were grouped in clusters ( Figs 5I–5L and S13A-S13F ) . By contrast , when the same components were knocked down in progenitor cells , the phenotype was either absent or very weak ( S13J–S13L Fig ) . In addition , ISC overproliferation was observed only when Wg signaling was disrupted during adulthood , but not prior to eclosion ( S13G–S13I and S13M–S13O Fig ) , consistent with our analysis of Wg pathway mutant clones . Together , these findings indicated that the activation of Wg signaling in enterocytes prevents the non-autonomous overproliferation of neighboring ISCs during adult homeostasis . We further sought to determine the mechanism by which loss of Wg signaling in enterocytes induces the non-autonomous overproliferation of nearby ISCs . We postulated that following Wg pathway inactivation , cytokines or ligands secreted from enterocytes could activate signaling , and thereby the proliferation of neighboring ISCs . This hypothesis is consistent with previous findings , which revealed that ligands from several different signal transduction pathways are released in this manner , including Unpaired ( Upd2 and Upd3 ) from the JAK-STAT pathway , Krn from the EGF pathway and Dpp from the TGF-β pathway [49–54] . Therefore , we compared the levels of transcripts encoding these ligands in control intestines with those in which signaling had been disrupted in enterocytes by RNAi-mediated knockdown of Wg pathway components or by expression of dominant negative Lgs or dominant negative TCF . We observed a marked increase in the expression of upd2 and upd3 , but none of the other ligands tested ( Figs 6A and S14A ) . Based on these results , we sought to determine whether the JAK-STAT pathway was aberrantly activated in ISCs following disruption of Wg signaling in enterocytes . Indeed , strong activation of stat-GFP expression , a JAK-STAT pathway reporter , was observed in clusters of wild-type progenitor cells near pygo null mutant clones ( Fig 6C–6C’ ) . In contrast , in cells farther away from mutant clones , or in cells adjacent to wild-type clones , stat-GFP was present at basal levels ( Fig 6B–6C’ ) . Consistent with this observation , strong induction of Socs36e , a direct target gene of the JAK-STAT pathway , was also detected following RNAi-mediated knockdown of Wg pathway components or overexpression of dominant negative Lgs ( S14B Fig ) . To determine whether JAK-STAT pathway activation mediates the non-autonomous effects of Wg pathway mutant enterocytes , we used RNAi-mediated knockdown to reduce upd2 and upd3 expression in enterocytes in which Wg signaling was disrupted concomitantly ( Fig 6D–6E ) . Notably , the aberrant increase in pH3+ cells and the abnormal clustering of Deltapos cells were both suppressed ( Fig 6D–6E ) . We conclude that JAK-STAT pathway activation is required for the non-autonomous overproliferation of ISCs that results from inhibition of Wg signaling in enterocytes . The JAK-STAT pathway is activated in response to various challenges in the midgut , including infection , apoptosis , and stress to promote rapid proliferation [50 , 55] . We sought to determine whether Wg pathway inhibition in enterocytes induces either a stress response and/or apoptosis , and secondarily results in JAK-STAT pathway activation and a proliferative response in neighboring ISCs . Therefore , we examined the activation of the two major stress-responsive signaling pathways , JNK ( c-Jun N-terminal kinase ) and Nrf2 ( Nuclear factor 2 ) [56–58] , through analysis of their respective target genes: puc and keap1 [59–61] . Importantly , we found that neither of these pathways was activated following Wg signaling disruption ( S14B Fig ) . Furthermore , enterocytes in which Wg signaling was inhibited using mutant clones or RNAi-mediated knockdown did not exhibit any hallmarks of apoptosis , including nuclear fragmentation , detachment from the epithelium , or caspase activation ( S14C–S14E’ Fig ) . These findings provided evidence that the induction of JAK-STAT pathway and ISC overproliferation is more likely a direct consequence of Wg pathway inactivation in these enterocytes . Together , we conclude that the maintenance of intestinal homeostasis requires activation of Wg signaling in enterocytes to prevent the non-autonomous activation of JAK-STAT signaling , and thereby the aberrant overproliferation of neighboring ISCs . The non-autonomous effect on ISC proliferation described above was observed within intestinal compartments . We also sought to determine the function of high-level Wg pathway activation at compartment boundaries and focused our analysis on the R5-HPZ border , which partitions the posterior terminal midgut ( R5 ) from the anterior hindgut [23 , 34] . This boundary is distinguished by the juxtaposition of two distinct epithelial cell populations that are derived from distinct origins and differ with respect to cell size , nuclear size and cell adhesion ( Fig 7A–7B’ ) [34 , 62] . To examine the roles of Wg signaling at the midgut-hindgut boundary , we generated mutant clones of the Wg pathway components arr , dsh , pygo and the functionally redundant fz and Dfz2 in larvae and analyzed adult guts shortly after eclosion . When wild-type clones crossed the midgut-hindgut boundary or were confined within the R5 region , normal cell morphology and cell-cell junctions were observed , and a discrete border between R5 and the HPZ was clearly demarcated , with high levels of Fas3 restricted to the hindgut ( Fig 7C–7D”‘ ) . In contrast , two distinct phenotypes were observed in Wg pathway mutant clones near the R5-HPZ boundary: the mutant epithelial cells either formed tightly-packed swirls or displayed markedly larger nuclear and cell size by comparison with their wild-type neighbors . The first class of mutant clones at the R5-HPZ boundary was comprised of masses of tightly-spaced cells of aberrant nuclear and cell size that were arranged in a spiral pattern , which were readily distinguishable from the surrounding loosely-ordered wild-type epithelium ( Figs 7E–7I”‘ and S15B–S15E”‘ ) . This “tightly packed” phenotype occurred with high penetrance ( S15A Fig ) and was often very severe . Of note , many of the tightly packed mutant clones extended outside the gut ( Figs 7F–7G”‘ , 7I–7I”‘ , S15B and S15E–S15E”‘ ) , and formed a hollow mass with a haze of DAPI staining at its center ( Figs 7F and S15B ) . Importantly , despite their location at the terminal midgut , these mutant clones expressed Fas3 at high levels , a characteristic that is normally restricted to the hindgut cells ( Figs 7F–7G”‘ and S15D–S15D”‘ ) . These observations suggested that the “tightly-packed” Wg pathway mutant clones failed to adopt a proper midgut fate , and instead displayed characteristics of the hindgut . To further test this conclusion , we analyzed the mutant clones for cell-specific markers . We found that the vast majority of mutant clones lacked Deltapos ISCs and Prosperopos enteroendocrine cells , and were also negative for Pdm-1 , a marker for differentiated enterocytes ( Fig 7H–7I”‘ ) [26 , 63] . In rare instances , Deltapos or Prosperopos cells were found at the clone periphery . In addition , no Cut was detected in the mutant clones , ruling out the possibility of misadoption of the renalcyte fate ( S15E–S15E”‘ Fig ) [64 , 65] . Together , these findings indicated that Wg signaling at the R5-HPZ boundary is critical for proper fate specification of posterior terminal midgut cells . The second major defect that we observed in Wg pathway mutant clones at the midgut-hindgut boundary were abnormally large cells with large nuclei as compared with their wild-type neighbors , which we termed “large cell” clones ( Fig 8A–8A”‘ ) . These large cell clones were found mainly in the midgut , but in some cases , they intercalated within the tightly-spaced rows of anterior hindgut cells in the HPZ ( S16A–S16A”‘ Fig ) . In sharp contrast with the “tightly-packed” clones , the cells within the large cell clones had low Fas3 levels and remained contiguous with the epithelial lining of the gut lumen ( S16B–S16B”‘ Fig ) . Thus , the “large cell” clones and the “tightly-packed” clones represented two distinct classes . Previous studies have shown that ISCs or EEs that underwent excessive replication and cell growth without cell division or differentiation displayed abnormally large nuclear and cell size [66 , 67] . To determine whether the “large cell” clones caused by inhibition of Wg signaling resulted from a similar mechanism , we stained these clones with midgut markers . Of note , Deltapos ISCs and Prosperopos EEs of normal size and ploidy were present inside these “large cell” clones; however , the cells of abnormally large size were negative for Delta and Prospero ( Figs 8B–8B”‘ and S16C–S16C”‘ ) . Further , many of the large cells were also devoid of Pdm-1 staining ( Fig 8C–8C”‘ ) , the loss of which correlated with the severity of the phenotype . Together , these findings indicated that the abnormally large cells had not adopted any of the known terminal cell fates of the midgut . Therefore , the “large cell” phenotype resulting from inactivation of Wg signaling was likely due to a defect in cell fate specification during midgut development . We conclude that Wg pathway activation ensures proper cell fate specification at compartment boundaries during the development of the adult intestine ( Fig 8D ) . The aberrant cell fate specification upon Wg pathway inhibition was not only observed at the midgut-hindgut boundary , but also at the foregut-midgut boundary , a site of high-level Wg pathway activity where the cardia forms [33 , 35] ( Fig 2B and 2B’ ) . Importantly , when wild-type clones crossed the cardia , neither nuclear morphology nor cell-cell junctions were affected ( S17A–S17A”‘ Fig ) . In contrast , several defects were observed in Wg pathway mutant clones located in the cardia . Specifically , normal cell alignment was disrupted , cells of abnormal size were detected , and further , mutant cells extended outside the cardia ( S17B–S17D”‘ Fig ) . Fz3-RFP was lost inside these mutant clones , confirming disruption of Wg pathway activity ( S17B–S17B”“Fig ) . Together , these observations provided evidence that Wg signaling directs proper cell fate specification at two major intestinal boundaries .
The adult fly gut is subdivided into distinct compartments that have unique function , histological structure , and gene expression repertoire [18–20] . Recently , the graded expression of fz3 , a known target gene of the Wg pathway , was found to peak at compartment boundaries [18] . This boundary-enriched activation of the Wg pathway could not be fully explained by the previously identified patterns of wg expression in the intestine [30 , 34 , 35] . In this study , we uncovered several novel regions of wg expression both in the gut epithelium and within the surrounding muscle . Together with previous reports , our new findings indicate that Wg is markedly enriched around intestinal compartment boundaries that coincide with prominent sphincters and tissue-organizing centers . An analogous enrichment of wg expression at major compartment boundaries was observed in the larval intestine . How these wg-enriched zones are established and maintained awaits further investigation . The Wnt pathway is known to be one of the key regulators of antero-posterior patterning and regional specificity in the development of vertebrate gastrointestinal tract [68 , 69] , suggesting that the requirement for Wg signaling at intestinal compartment boundaries in flies might be evolutionarily conserved . Previous studies of Wg signaling in intestinal homeostasis were focused on transduction of the pathway in ISCs [25 , 26 , 30 , 70] . Here , however , starting with Wg signaling reporters that enabled cell-type specific analysis and coupling these findings with functional studies , we have discovered that the primary site of Wg pathway activation during adult homeostasis is in enterocytes and not ISCs . Two factors may account for the discrepancy between our results and previous reports . First , the primary conclusions regarding strong effects on ISC self-renewal in previous studies were based mainly on a dominant negative TCF ( dnTCF ) [25 , 26 , 30] , which is a truncated and overexpressed protein [4] . Indeed , the strong negative effects on ISC proliferation resulting from dnTCF expression were not recapitulated by null alleles of the essential components Frizzled or Pygo [25] , indicating that dnTCF exhibited some phenotypes that were not present upon complete inactivation of Wg pathway , and thus some dnTCF effects do not represent the physiological roles of Wg signaling in intestinal homeostasis . For this reason , we based our studies on null alleles , and tested our conclusions by analyzing multiple essential Wg pathway components . Second , the initial study investigating the roles of Wg signaling on ISC renewal examined mutant clones 30 days after clone induction [30] . Our clonal analysis was performed at much earlier timepoints ( 5–7 days ACI ) . Therefore , the differing results may indicate that Wg signaling has biphasic roles in adult gut homeostasis . Of note , we found that all cell types in the adult intestinal epithelium have the capability to respond to Wg exposure . Therefore , the responsivity of different intestinal cell types to Wg stimulation may reflect inherent differences in their threshold for pathway activation . We postulate that the threshold for Wg pathway activation is higher in ISCs than enterocytes , and therefore signaling is activated in ISCs only under hyperactivated contexts , as in Apc mutants [26–29] , or in response to high levels of wg that are expressed following intestinal injury [25] . Nonetheless , ISCs at the R5-HPZ boundary are an exceptional population in which the Wg pathway is active during homeostasis . The R5-HPZ compartment boundary is a unique region at which three distinct sources of the Wg ligand converge , and exhibits the highest level of Wg pathway activation in the gut ( our observations and [18] ) . Therefore , the level of wg in this region may surpass the threshold required for activation of signaling in ISCs . We have found that when Wg signaling is inactivated in enterocytes , the JAK-STAT pathway is aberrantly induced in neighboring wild-type ISCs , and drives their non-autonomous proliferation ( Fig 8D ) . In wild-type flies , the JAK-STAT pathway is activated in response to various challenges such as infection , apoptosis , and stress to promote rapid ISC proliferation [50 , 55] . However , we found that neither of the two major stress response pathways in the fly gut , JNK and Nrf2 , nor apoptosis , was induced following Wg pathway inhibition in enterocytes . Therefore , the non-autonomous ISC overproliferation is not a secondary consequence of stress or cell death in neighboring enterocytes . Transcription factors including TCF that are critical for maintenance of gut regionalization in the adulthood can also regulate homeostasis [18] . Therefore , inactivation of Wg signaling during adult homeostasis could potentially disrupt gut regionalization and thus trigger JAK-STAT pathway activation . Alternatively , Wg pathway activation might be required for the tight control of JAK-STAT signaling during homeostasis . As a critical barrier to toxins and infection , intestines are subjected to constant injury and activate JAK-STAT signaling as a compensatory response . We postulate that Wg signaling is required to prevent the inappropriate activation of this critical response during homeostasis and that this “brake” must be shut off or bypassed during regeneration following injury . Wg pathway activation has known critical roles in cell sorting and patterning at several distinct compartment boundaries in metazoans , including at fly embryonic parasegmental boundaries , the larval wing disc dorsal-ventral boundary , and vertebrate rhombomere boundaries [71–74] . Here , we have found that Wg pathway activation may serve a similar tissue-organizing role at intestinal compartment boundaries , where it is required for proper cell fate specification and lineage separation . We focused on the R5-HPZ boundary , a site of high-level Wg pathway activity , where cells of completely distinct functional and morphological properties are separated [34] . During development of the adult intestine , the R5 compartment is formed by the anterior migration of HPZ cells from the hindgut to the midgut accompanied by their re-specification as midgut cells , and the posterior migration of midgut AMPs [62] . Despite this striking bi-directional movement , the precise coordination of cells in this region ensures proper fate specification and demarcation by a sharp border . Here , we discovered that disruption of Wg signaling at the R5-HPZ boundary results in two distinct defects: the formation of either “tightly packed” groups of cells or abnormally “large” cells ( Fig 8D ) . Both of these defects affect proper cell fate specification and boundary maintenance , but are disparate in nature . First , the tightly packed cells segregate away from their wild-type neighbors , and cluster together to form a hollow spherical mass that can grow outside the midgut . Intriguingly , despite their location in the midgut , these “tightly packed” cells lack cell-type specific midgut markers , and express Fas3 at high levels that are normally found in the hindgut cells . Based on these observations , we speculate that these “tightly packed” clones are hindgut cells that fail to adopt midgut cell fate following migration . Unlike the “tightly packed” cells , the abnormally “large” cells that also result from inactivation of Wg signaling are likely midgut-derived , as they express Fas3 at low levels and are contiguous with midgut epithelium . Previous reports revealed that defects in mitosis result in dysfunctional ISCs that undergo replication without division [66 , 67] , which in principle could have been the cause of the aberrantly large cells we observed . Similarly , mitotic deregulation in EEs also results in an abnormal increase in ploidy [67] . However , the “large cells” that resulted from Wg pathway inhibition are devoid of normal ISC and EE markers . Further , almost all of the abnormally large cells do not express markers characteristic of differentiated enteroctyes . Importantly , these “large cell” clones are multi-cellular , and include ISCs and EEs of normal ploidy , size , and cell-type specific markers . These characteristics distinguish the phenotype of the Wg pathway mutant clones from the previously reported large cells that result from defective mitosis [66 , 67] and suggest that these large cells , like the “tightly-packed” cells , likely arise from deregulated cell fate specification and have likely adopted an intermediate or novel cell fate .
Crosses were performed at 25°C except those under control of gal4/gal80ts driver , which were set up at room temperature . To test the effects specifically during adulthood , progeny were collected within 2 days after eclosion and shifted to 29°C for another 7–10 days . To determine the potential function during development , 2nd instar larvae were switched to 29°C and dissected right after eclosion . For the rpr-induced cell death experiments , flies overexpressing rpr , as well as control flies , were shifted to 29°C 3 days after eclosion for 40 hours before examination . Primary antibodies used for immunostaining were mouse anti-β-gal ( Promega ) 1:500 , rabbit anti-β-gal 1:5000 ( MP Biomedicals ) , mouse anti-Arm 1:20 ( [83]; Developmental Studies Hybridoma Bank , DSHB ) , mouse anti-delta 1:100 ( [78]; DSHB ) , mouse anti-Prospero 1:20 ( [84]; DSHB ) , mouse anti-Cut 1:20 ( [85]; DSHB ) , mouse anti-α-Spectrin 1:20 ( [86]; DSHB ) , mouse anti-Fas3 1:20 ( [87]; DSHB ) , rabbit anti-Pdm-1 1:200 [63] , rabbit anti phosphor-histone H3 ( Ser10 ) 1:1000 ( Millipore ) , rabbit anti-DsRed 1:500 ( Clontech ) , chicken anti-GFP 1:10000 ( Abcam ) , rabbit anti-GFP 1:500 ( Life Technologies ) , Alexa Fluor 555 phalloidin 1:500 ( Life Technologies ) , Alexa Fluor 488 phalloidin 1:500 ( Life Technologies ) , rabbit anti-cleaved Drosophila Dcp-1 ( Asp216 ) 1:100 ( Cell Signaling ) and DAPI 1:100 ( Sigma ) . Secondary antibodies used were goat or donkey Alexa Fluor 488 or 555 conjugates at 1:400 ( Life Technologies ) , and goat or donkey Cy5 conjugates at 1:200 ( Life Technologies/Jackson Immunochemicals ) . Confocal images were captured on a Nikon A1RSi confocal microscope and processed with Adobe Photoshop software . Adult fly guts were dissected in PBS , fixed in 4% paraformaldehyde for 45 mins to 1hr and washed with PBS+0 . 1% Triton X-100 . Specifically , adult wg-lacZ guts were fixed in sodium cacodylate buffer [48] for 25 mins , and for Delta antibody staining , guts of desired genotypes were fixed in sodium cacodylate buffer for 20 mins . After blocking with PBS+0 . 1% Tween-20+10% BSA for 1h at room temperature , the samples were incubated with primary antibody ( diluted in PBS+0 . 5% Triton X-100 ) at 4°C overnight . Secondary antibody incubation was carried out at room temperature for 2 hrs . The samples were subsequently stained with DAPI ( 2 μg/ml ) and mounted in Prolong Gold Antifade Reagent ( Life Technologies ) . Larval guts were immunostained in the same way except that the wandering 3rd instar larvae were dissected and fixed for only 20 mins . More than 20 guts of desired genotypes were examined unless specified and the representative image is shown . The non-autonomous ISC overproliferation defects were observed in both anterior and posterior midguts , and the specific subregions shown in the figures are indicated in the figure legends . Quantification of fz3-RFP and nkd-lacZ expression level in posterior midgut was performed by NIS-Elements software . Stat-GFP intensity was analyzed via Imaris software ( Bitplane ) . For ISC quantification , flies were stained with anti-Delta and anti-Prospero antibodies . 60x images of posterior midgut region ( R4-R5 ) were obtained and the total number of Delta positive cells in the field was counted . T-tests were performed using Prism ( GraphPad software , USA ) . MARCM clones were generated as described [40 , 41] . To induce clones during development , 1st and 2nd instar larvae were heat shocked in a 37°C water bath for 2–3 hrs , except for crosses driven by one MARCM 82B driver [79] , which were induced during larval-pupal transition . For specific MARCM lines , the heat-shock was repeated the next day . To generate MARCM clones during adulthood , progeny of the desired genotype were collected within 2 days after eclosion and were heat-shocked in a 37°C water bath . The length of the heat shock varied with the MARCM line , from one 30-min heat shock to four 90-min heat shocks over 2 days . The adult clones were examined 7–10 days later . For larval clones , 1st and 2nd instar larvae were heat-shocked for 2–3 hrs and guts of wandering 3rd instar larvae were examined . 15–20 fly guts of proper genotype and age were dissected and homogenized in Trizol ( Invitrogen ) . Total RNA was extracted according to the protocol from the Drosophila Genomics Resource Center or using the RNA miniprep kit ( Zymo research ) . The RNA was subsequently treated with RQ1 DNase ( Promega ) . 1 μg of RNA was reverse transcribed using p ( dT ) 15 primers ( Roche ) and M-MLV reverse transcriptase ( Invitrogen ) . Expression level of candidate ligands was quantified using the StepOne Real-time PCR system ( Life Technologies ) with SYBR green ( Life Technologies/Biorad ) and the results were presented as mean fold change with standard deviation . The primers for rpl32 ( internal control ) , upd3 and Socs36e were adapted from [28] , whereas primers for krn were adapted from [88] . The other primer sequences were: dpp F: 5’- TCT GCT GAC CAA GTC GG -3’ , dpp R: 5’- GCG GGA ATG CTC TTC AC -3’ , upd2 F: 5’- TGG TAT TCG CTC ATC GTG A -3’ , upd2 R: 5’- GGC AAA TCA GAG ATC CCG -3’ . puc F: 5’- CAC ATC AGA ACA TCA AGC AGT AC –3’ , puc R: 5’-GTA GGC GAT GGC AAT GG -3’ and Keap1 F: 5’-TAC AAG AGT CCA GCG ATC CA –3’ and keap1 R: 5’-GTC ACC GAA ACA TGG CGT-3’ . | The highly conserved Wingless/Wnt signal transduction pathway directs many cellular processes in metazoans and its deregulation underlies numerous human congenital diseases and cancers . Most notably , more than 80% of colon cancers arise from aberrant activation of the Wnt pathway . A better understanding of how Wnt signaling functions in the intestinal stem cells ( ISCs ) during homeostasis and in disease states is thus critical . The Drosophila digestive tract provides a powerful genetic model and an entry point to study these questions . Here , we find that the Wg ligand and pathway activation are enriched at Drosophila intestinal compartment boundaries and are essential for development and homeostasis of the adult gut . During homeostasis , Wg signaling in enterocytes is required to prevent the overproliferation of ISCs non-autonomously . In addition , during development , Wg signaling ensures proper cell fate specification near compartment boundaries . These findings provide insight into the mechanisms underlying the Wg-dependent regulation of adult intestinal function . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"medicine",
"and",
"health",
"sciences",
"cloning",
"physiological",
"processes",
"age",
"groups",
"adults",
"homeostasis",
"stem",
"cells",
"molecular",
"biology",
"techniques",
"cardia",
"digestive",
"system",
"research",
"and",
"analysis",
"methods",
"animal",
"cells",
"biological",
"tissue",
"molecular",
"biology",
"jak-stat",
"signaling",
"cascade",
"gastrointestinal",
"tract",
"people",
"and",
"places",
"signal",
"transduction",
"anatomy",
"cell",
"biology",
"physiology",
"epithelium",
"biology",
"and",
"life",
"sciences",
"population",
"groupings",
"cellular",
"types",
"cell",
"signaling",
"signaling",
"cascades"
] | 2016 | Regulation of Stem Cell Proliferation and Cell Fate Specification by Wingless/Wnt Signaling Gradients Enriched at Adult Intestinal Compartment Boundaries |
Mycoplasma pneumoniae , a human pathogenic bacterium , glides on host cell surfaces by a unique and unknown mechanism . It forms an attachment organelle at a cell pole as a membrane protrusion composed of surface and internal structures , with a highly organized architecture . In the present study , we succeeded in isolating the internal structure of the organelle by sucrose-gradient centrifugation . The negative-staining electron microscopy clarified the details and dimensions of the internal structure , which is composed of terminal button , paired plates , and bowl complex from the end of cell front . Peptide mass fingerprinting of the structure suggested 25 novel components for the organelle , and 3 of them were suggested for their involvement in the structure through their subcellular localization determined by enhanced yellow fluorescent protein ( EYFP ) tagging . Thirteen component proteins including the previously reported ones were mapped on the organelle systematically for the first time , in nanometer order by EYFP tagging and immunoelectron microscopy . Two , three , and six specific proteins localized specifically to the terminal button , the paired plates , and the bowl , respectively and interestingly , HMW2 molecules were aligned parallel to form the plate . The integration of these results gave the whole image of the organelle and allowed us to discuss possible gliding mechanisms .
Mycoplasmas are parasitic and occasionally commensal bacteria that lack a peptidoglycan layer and have small genomes [1] . Mycoplasma pneumoniae is a causative pathogen of human bronchitis and walking pneumonia . Outbreaks of mycoplasma pneumonia occur frequently in many parts of the world , and the increase of macrolide-resistant M . pneumoniae is a growing problem [2–4] . M . pneumoniae exhibits gliding motility in the direction of the protrusion at a maximum speed of 1 μm , one-half its cell length , per second [5–8] . This motility , combined with the ability to adhere to epithelial cells , is involved in the pathogenic process , enabling the cells to translocate from the tips of bronchial cilia to the host cell surface [9] . Previous studies , including genome analyses , have shown that this motility is not related to other known mechanisms of bacterial motility , nor does it involve motor proteins known to be involved in eukaryotic cell motility [5 , 10–14] . The gliding machinery is a membrane protrusion formed at one pole and can be called an attachment organelle , which is composed of a naplike surface structure and an internal core [15 , 16] . The naplike structure corresponds to the complex of P1 adhesin , a plausible leg and also a receptor of sialylated oligosaccharides , which are the ligands on host tissue surfaces [17–19] . The other structure , the internal core , has been known for more than 40 years as a structure that shows rather high contrast in images [20 , 21] . Recent studies suggested that this structure can be divided into three parts , including a terminal button , thick and thin paired plates , and a bowl ( wheel ) complex from the front end [11 , 15 , 16 , 22] , and that it is essential for the formation of an attachment organelle but does not include any conventional cytoskeletal proteins such as MreB or FtsZ , unlike other bacterial structures [11 , 13] . This structure should be rather stable because it can be isolated on an electron microscopy ( EM ) grid when the cells are extracted by Triton X-100 [20–23] . However , both detailed and whole images of the internal core are still unclear . Although protein localizations have been shown for four surface proteins , P1 adhesin , P40 , P90 , and P30 , as well as for eight internal proteins , P65 , HMW2 , P41 , HMW1 , HMW3 , P200 , TopJ , and P24 , they have not been mapped systematically on the cell images [5 , 24] . In the present study , we succeeded in isolating the internal core , observed the details , identified novel protein components , mapped the component proteins of the organelle by protein tagging , and then suggested images for the architecture and the gliding mechanism .
To visualize the core in detail and to elucidate the component proteins , we isolated a fraction rich in internal cores . M . pneumoniae cells recovered from a tissue culture flask were suspended in a buffer , treated with different concentrations of Triton X-100 or Tween 20 for 3 min , and decided the suitable conditions for isolation ( Fig 1A , 1B and 1C ) . The cell fraction treated by 1% Triton X-100 was subjected to stepwise gradient centrifugation , consisting of 0% , 20% , 30% , 40% , 50% , and 60% sucrose layers . After centrifugation , we found a dense layer at the bottoms of the 40% sucrose layers . We recovered and observed the fraction under EM and found that the fraction visually contained only the internal cores , whose features were common to those isolated on a grid , with similar appearances and dimensions , but which did not show other structures such as membrane pieces under EM ( Fig 1D and 1E ) . The images of internal cores allowed us to categorize the cores into four types ( Fig 2 and S1 Fig ) . The "bold" type featured a wide and straight central part ( Fig 2A ) . The "slim" type featured a narrow and bent central part ( Fig 2B ) . The "fork" type featured a lucent part near the center ( Fig 2C ) . The "branched" type featured a branch near the back end of core ( Fig 2D ) . The ratios of these images in the microscopic field were 51% , 39% , 4% , and 6% , respectively . The average lengths of the bold , slim , and fork types were 316 ± 16 ( n = 40 ) , 291 ± 11 ( n = 40 ) , and 309 ± 18 ( n = 10 ) nm , respectively . As previously suggested , all of the images could be divided into three parts: a terminal button at the front end , central paired plates , and a bowl ( wheel ) complex at the back end [11 , 15 , 16 , 22] . Both the bowl complex and the terminal button could be divided into two parts ( Fig 2E and 2G ) . The backside part of bowl complex has not been visualized in previous studies [15 , 16 , 20 , 21] . Although the structures at both ends were common among all four types , the central "plate" was specific to each type . In the bold type , the plate was rather symmetrical around its axis , with a maximum width of 60 ± 10 nm ( n = 40 ) . In the slim type , the narrow plate had a bend around the position at 40–60 nm from the back end ( Fig 2H ) . The plate of the fork featured a low-density area near the center . In the branched type , the plate was branched near the bowl complex with duplicated terminal buttons . In the isolated core fractions observed here , the detailed structures were visible more clearly than the structures of the cells treated on EM grids as shown in Fig 1 , probably because the structures covering the internal cores were removed in the isolated fraction . The proteins involved in the core fraction were identified by peptide mass fingerprinting ( PMF ) ( Fig 3 ) using the following procedure . ( i ) The proteins of the whole cell lysate , soluble , and core fractions were developed by SDS-PAGE and applied to Coomassie Brilliant Blue ( CBB ) staining , in which the band density is better related to the protein amount than other staining methods [25] . ( ii ) Thirty-four bands isolated from the lane of the core fraction were identified by PMF . ( iii ) To identify the proteins specific to the core fraction , the protein bands detected at the same position in the lane of the soluble fraction were also identified by PMF . If these peptides were derived from different proteins or if no bands were found at the corresponding position in the soluble fraction , the protein from the band was defined as specific to the core fraction . ( iv ) To examine the possibility that an apparently single band contains more than one proteins , the Mass Spectrometry signals which were not assigned to the identified proteins were subjected to the next identification . Finally , we identified 34 proteins from the core fraction and listed 37 proteins as components of the attachment organelle and candidates in four categories , as follows ( S1 Table ) . ( Category I ) Nine proteins identified in the present study were already known as the components of the attachment organelle: P1 adhesin , P90 , P40 , P65 , P41 , HMW1 , HMW2 , HMW3 , and P200 [5 , 26] . P1 adhesin , P90 , and P40 were found also in the soluble fraction , suggesting that a fraction of these proteins might be solubilized by the detergent because they localized on the cell surface [18] . ( Category II ) Three proteins , TopJ , P24 , and P30 , were not found in the core fraction , although they have been reported as the organelle components [5] . These proteins may detach from the attachment organelle easily , or they may exist in smaller amounts than the other component proteins . ( Category III ) Six proteins were found specifically in the core fraction . ( Category IV ) Nineteen proteins were found in both the core and soluble fractions . The molar ratios of proteins specific to the core fraction were determined by band densitometry . The previously reported component proteins were found in higher protein amount than the others . We listed 37 proteins in S1 Table , composed of 12 proteins previously reported as a component of the attachment organelle ( categories I and II ) and 25 proteins found in the core fraction in the present study ( categories III and IV ) . Six proteins in category IV , PtsG ( MPN207 ) , OppF ( MPN218 ) , PdhB ( MPN392 ) , PdhA ( MPN393 ) , GlyA ( MPN576 ) , and Tuf ( MPN665 ) , have been known to be a part of large complexes involved in cell metabolism , which likely remained in the core fraction [27] . Therefore , we focused on the other 19 proteins in categories III and IV as candidates for novel components of the attachment organelle . To examine the involvement of these 19 proteins in the attachment organelle , we applied EYFP fusion tagging to 31 proteins including categories I and II ( S2A Fig ) [7 , 28] . In this system , the M . pneumoniae genes of interest were fused by the eyfp gene at their N-termini by using the Gateway subcloning method , as described previously [7] . The fused genes were delivered into random positions on the M . pneumoniae chromosome using the Tn4001 transposon vector system [29] , and expressed in the wild-type background by a strong promoter of the tuf gene [7] . First , we applied this strategy to all 12 proteins in categories I and II , which are known components . As shown in the legend of S2 Fig , the localizing patterns of the EYFP signal were divided into four types: type "a" had a focused signal at the attachment organelle , type "f" had focused signals at different positions than the organelle , type "d" had a diffuse signal in the whole cell , and type "n" had no signal . The EYFP fusions with P65 ( MPN309 ) , HMW2 ( MPN310 ) , P41 ( MPN311 ) , HMW1 ( MPN447 ) , HMW3 ( MPN452 ) , and P24 ( MPN312 ) showed clear localization at the organelle ( type "a" ) , consistent with previous reports [7 , 12 , 30–32] . P200 ( MPN567 ) and TopJ ( MPN119 ) showed localization at the organelle ( type "a" ) , but also additional localization elsewhere ( type "f" ) . Probably , this was caused by the excess of proteins produced from the strong promoter , tuf , stacked in a space in a cell usually at the other end of the organelle . The expression of any EYFP fusions did not show obvious defects in the binding or gliding activities of M . pneumoniae . At room temperature ( RT ) , less than 1 . 3% of cells ( n = 150 ) attached to the glass bottom dish glided with gliding speeds less than 0 . 01 μm/s , 40 times slower than those at 37°C . We then took pictures at RT without fixation . P1 adhesin ( MPN141 ) and P30 ( MPN453 ) showed diffuse signals ( type "d" ) , although they are known to localize at the attachment organelle as shown by immunofluorescence microscopy [12 , 31 , 32] . Probably , this was caused by the steric effect of EYFP fusion to the N terminus on protein folding or sorting . We next examined the C-terminus fusion and found that the signals of P1 adhesin and P30 were focused at the attachment organelle , suggesting that the fusion site is critical in some cases [7 , 28 , 30 , 33 , 34] . MPN142 is known to be cleaved into two proteins , P40 and P90 , which are composed respectively of 1–454 and 455–1218 amino acid residues after translation and which localize at the attachment organelle [12 , 31 , 32 , 35] . We tried EYFP fusion to both sides of MPN142 and to the N terminus of the P90 sequences . The N-terminal tagging of MPN142 showed diffuse subcellular localization , and SDS-PAGE showed that the cleavage of MPN142 did not occur . The C-terminal tagging of MPN142 and the N-terminal tagging of the P90 part of MPN142 showed no and diffuse fluorescence in a cell , respectively . Next we examined the subcellular localization of the 19 candidate proteins in categories III and IV by N-terminal fusion and found that 2 proteins , MPN387 and MPN332 , localized at the attachment organelle ( Fig 3B ) . MPN387 , found specifically in the core fraction , is featured in its sequence by a coiled-coil domain spanning 87–285 amino acid residues in the total of 358 . MPN332 , found in both the core and soluble fractions , has been annotated as Lon protease , an ATP-dependent protease [13] . C-terminal fusion was also applied to the proteins found specifically in the core fraction , including MPN066 , MPN140 , MPN295 , MPN372 , and MPN627 , revealing that MPN066 , annotated as "CpsG" , phosphomannomutase , or phosphoglucomutase [13] , localizes at the organelle . On the basis of these results , we concluded that the attachment organelle includes three additional novel components: MPN066 , MPN387 , and MPN332 . MPN332 was detected also in the soluble fraction , suggesting that this protein departs from the attachment organelle easily or exists abundantly . In the PMF , four proteins were identified specifically in the core fraction but were not proven to localize at the attachment organelle by EYFP fusion tagging:MPN140 ( ORF4 ) , MPN295 , MPN372 ( CARDS toxin ) , and MPN627 ( PtsI ) . Namely , the fluorescence images of MPN140 , MPN372 , and MPN627 showed diffuse patterns , and the signals of MPN295 focused on positions other than the attachment organelle . These proteins may be involved in other protein complexes that behaved similarly to the internal core in the fractionation by centrifugation . Otherwise their localization to the attachment organelle may be affected by the EYFP fusion . To address the possibility that additional focal localization was caused by heterogeneity in the expressed fusion proteins , we detected the fusion proteins for P200 ( MPN567 ) , TopJ ( MPN119 ) , MPN295 , MPN390 ( PdhD ) , and MPN391 ( PdhC ) by Western blotting using an antibody against EYFP ( S3 Fig ) . The major bands were detected around the positions expected from the amino acid sequences . These observations suggest that the focal localization at positions other than the organelle was not caused by the heterogeneity in expressed proteins . To obtain a clear image of the attachment organelle , we tried to map the fluorescent foci of the 13 component proteins tagged by EYFP , on an attachment organelle that was about 450 nm long ( Fig 4 ) . The fluorescence intensity of a cell with a typical extended shape was profiled along the cell axis and its peak position was determined ( S4 Fig ) . The distribution range depended on the proteins , i . e . P1 adhesin spread more than P30 , HMW3 , HMW1 , and P65 , as reported previously [31 , 32] . To map the fluorescent foci on the cell axis precisely , the phase-contrast image was used to determine the front end position of a cell , assuming that the attachment organelle is uniform in shape and density among individual cells ( S5 Fig ) . The cell image density was normalized by a control object , a 200-nm polystyrene bead as the 100% density , because the image density depends on the illuminating conditions generally in bright-field optical microscopy . As P30 , a transmembrane protein [36] was found to localize at the front side , the position of this protein along the cell axis corresponding to 16 ± 5% cell image density ( n = 60 ) was set as the zero position . Therefore , we mapped the axial peak positions of fluorescence as the distance from the position where the image density of cell was 16% of that at the bead center . This process was applied to 20 cells , each labeled for the 13 proteins , and the average positions were mapped ( Fig 4A ) . Note that these histograms show the variations in peak positions of cells , not the signal distributions in a single cell which is shown in S4 Fig . The analysis of the amino acid sequence of HMW2 suggests that 11 regions , composed of 1257 amino acid residues of the whole 1818 residues , form a coiled-coil structure [37] . There have been conflicting data regarding the alignment of HMW2 molecule in the internal core , namely parallel [37 , 38] or perpendicular [39] to the core axis . If the molecules are aligned parallel , the total length is predicted to be 190 nm . Therefore , we fused EYFP to the C-terminus of this gene and analyzed the signal positions . The signal labeled at the C-terminus was localized at 257 ± 56 nm from the front , and the signal at the N-terminus was localized at 55 ± 60 nm , suggesting that HMW2 molecules are aligned in parallel along the cell axis , pointing the N terminus to the front ( Fig 4B and 4C ) . The 93% and 95% of isolated fluorescent foci were found in the organelle for N-terminus and C-terminus fusions , respectively ( S3 Fig ) . The Western blotting analyses showed the major bands at the positions close to those of 245 and 249 kDa expected from the amino acid sequences , respectively for the N-terminal and C-terminal fusions . The internal translation product corresponding to amino acids 1620–1818 of HMW2 , designated as P28 was also detected near the expected gel position with the relative protein amount to HMW2 consistent with the previous study [40 , 41] . These results suggest that the fusion proteins were expressed and incorporated into the organelle in manners similar to those of the original HMW2 and P28 , because the signals were not detected in additional positions . If this is the case , the part of HMW2 corresponding to the P28 may be responsible for the localization of the C-teminal part . All sets of data pairs were examined by ANOVA test ( S3 Table ) and the proteins were classified into six clusters as summarized in S6 Fig . Next , the relative positions of 16 pairs of proteins were also examined , by measuring the distances between fluorescent foci labeled by different fluorophores , including EYFP , ECFP , and Cy3 ( Fig 4 and S2 Fig ) . The distances between the positions within all 16 pairs were in good agreement with the values expected from the positions of the individual proteins , showing that the mapping of individual proteins using the density of phase-contrast image is reliable . To map those positions onto EM images , immunogold-EM was applied . The cells labeled by EYFP fusion for 13 protein constructs were attached to an EM grid , treated with Triton X-100 to expose the internal core , and then labeled with antiserum against EYFP conjugated by 5 nm of colloidal gold ( Fig 5 ) . P1 adhesin was not examined because it is anchored to the cell membrane which is removed by the detergent treatment [18] . The images of internal cores were less clear than those observed in Figs 1 and 2 , and S1 Fig , plausibly due to the labeling process by the antibodies . Specific labeling by gold was detected for six constructs , P65 , HMW2 ( N-terminus fusion ) , HMW3 , HMW1 , P200 , and P24 , around the internal core , but the other seven constructs , HMW2 ( C-terminus fusion ) , P41 , TopJ , P30 , CpsG ( MPN066 ) , MPN387 , and Lon ( MPN332 ) did not show specific gold labeling . The structure and assembly of the proteins successfully detected in this manner may be suitable to be exposed by the Triton X-100 extraction , because the fused EYFP was detected by the antibody . On the other hand , the proteins unsuccessfully detected in this manner may be removed by the detergent or they may not be exposed even after the detergent treatment . The gold positions were plotted along the long axis , fitted by Gaussian distribution , and mapped along the cell axis ( Fig 5 ) . Most of those positions were in good agreements with the fluorescence images ( Fig 6A ) . The position of HMW3 did not agree well with that from fluorescence , suggesting that this protein was slightly detached from the core by the Triton treatment for immuno electron microscopy . The distribution of gold labeling showed the following localization: P65 , terminal button; HMW2 ( N-terminus ) , front of paired plates; HMW3 , front side of paired plates; HMW1 , paired plates; P200 and P24 , bowl complex .
In the present study , the internal core was fractionated through Triton X-100 treatment and centrifugation , and then observed by EM ( Fig 1 ) . This procedure gave us images with higher contrast and lower backgrounds than those in previous studies without fractionation [6 , 15 , 16 , 20 , 21] . In a previous study , a similar approach was applied to the internal core of M . pneumoniae [23] . However , the sucrose-gradient centrifugation applied in the present study allowed us to isolate internal cores with less damages to the structure than occurred in the previous study ( Fig 2 and S1 Fig ) . In the present images , the isolated core structure included the bowl complex , which could be observed only in the electron cryotomography ( ECT ) of M . pneumoniae , and negative staining of Triton extracted M . gallisepticum , a species related to M . pneumoniae [6 , 15 , 16] . The images obtained here were clear enough to determine the detailed dimensions for the first time [15 , 16 , 20 , 21] , but not uniform enough for image averaging for higher resolution , suggesting that the internal core has some flexible parts . This feature is shared with the internal structure of M . gallisepticum [6] . How are the four types of images obtained in the present study related to one another ( Fig 2 and S1 Fig ) ? Previous analyses by fluorescence microscopy showed that the attachment organelle replicates for cell reproduction[24 , 30 , 32] . On the basis of this observation , the possible relationships among these four types can be traced from the slim type , which is probably a side view of an internal core in a cell ( Fig 6B ) . Possibly , the fork type may be a partly duplicated form of the slim type , where the daughter plates are two-storied . The branched type may be the later stage of replication . The attachment organelle has been suggested to replicate prior to cell division , based on the duplication scheme of the fluorescence signals labeled to the component proteins , and the frequency of branched type observed in the present study ( 4% ) is consistent with the previous observations showing that 5 . 5% of growing cells have two signals of the attachment organelle at adjacent positions [30 , 32] . The bold type is the top or bottom view of these three forms . In the internal structure of M . gallisepticum extracted from a cell by Triton X-100 also , an image similar to the branched type was observed with a frequency comparable to the frequency of cells with two organelles under optical microscopy [6] , suggesting that the replicating scheme is conserved in the related species . However , additional experiments are needed to conclude this reproduction scheme , because the fork and branched structure could be converted from the bold type in the core isolation process . Previous ECT imaging of the whole cell showed that the middle part of the internal structure is composed of thick and thin plates [15 , 16] . In the present study , however , the plate was not resolved into two layers in the core fraction ( Figs 1 and 2 ) , suggesting that the thin plate was removed in Triton treatment or the paired plates stick together in the dried condition of negative staining . Similarly , the distal ends of terminal button and bowl appeared differently from those in ECT [15 , 16] , namely some accessories were observed in the present study ( Fig 2 ) . Although the accessories did not show uniform shapes , they were observed reproducibly . Probably , these structures were exposed and enhanced for contrast by the detergent extraction and the staining by molybdate . We should consider that the negative staining EM after extraction and isolation may tend to give images with higher contrast and more changes from the original than those from ECT . In a previous study , M . pneumoniae cells were extracted by Triton X-100 and analyzed by mass spectrometry to identify the protein components of the insoluble fraction [23] . Our present study adopted a similar strategy but contains three critical improvements , as follows ( Fig 3 ) . ( i ) The proteins in the soluble fraction were also identified as reference . ( ii ) The analyzed fraction was well isolated from others and did not contain other structures based upon examination by EM . ( iii ) Identified proteins were further analyzed by EYFP tagging . In the present study , 25 proteins were suggested as candidates for novel core components by PMF , and then MPN066 , MPN332 , and MPN387 were suggested to be components by EYFP tagging ( Fig 3C ) . MPN387 is conserved among the genomes of related species , M . genitalium and M . gallisepticum [42 , 43] . This protein has been suggested to be involved in the gliding mechanism , because the depletion of this protein by a transposon insertion induced a nongliding phenotype , retaining hemadsorption activity [44] . MPN387 localized around the bowl complex ( Figs 3B and 4 ) , in common with P200 , which is involved in movement rather than binding [45] . These findings suggest that the bowl complex has a critical role in force generation for gliding . The other two proteins , MPN066 and MPN332 , have annotations to proteins widely distributed in bacterial genomes [13] . MPN066 is CpsG , a phosphomannomutase or phosphoglucomutase involved in sugar metabolism [46] . MPN332 is a Lon protease which have a role in diverse cellular functions in bacteria [47 , 48] . The roles of these proteins in cytadherence and gliding are unclear . It has been suggested that the attachment organelle may have roles in addition to adhesion and gliding , because usually the replication of this organelle occurs as an early event in the cell division process [30 , 32] . We found three proteins as novel components of the attachment organelle , on the basis of the EYFP tagging results . However , EYFP fusion sometimes has steric effects on the folding and sorting of the proteins of interest [7 , 28] . Therefore , the involvements of other candidates to the attachment organelle cannot be ruled out , especially for the proteins in category III . Now , 15 proteins have been assigned as components of the attachment organelle , including three newly identified proteins . We then systematically mapped 13 proteins that were available for EYFP tagging , and six of those 13 proteins also on EM images , and suggested an integrated image for the attachment organelle which is consistent with previous studies , as follows ( Fig 6A and 6C ) . ( i ) The terminal button is composed of P65 and HMW3 [31 , 32] . HMW3 localize at the boundary with the paired plate . ( ii ) The paired plates are composed of HMW1 , HMW2 , and CpsG [31 , 32 , 38] . HMW2 molecules are aligned in parallel along the axis with the N terminus at the front extending into the terminal button and the bowl complex [37 , 38] . HMW1 localizes at distant positions from the paired plates as observed in the slim core EM images ( Fig 5 ) , which is consistent with the "peripheral association" which was reported previously [49] . ( iii ) The bowl complex is composed of MPN387 , P41 , P200 , TopJ , P24 , and Lon protease [7 , 45 , 50] . MPN387 and P200 are known to be necessary for gliding rather than binding [30 , 45] . P41 has a role in connecting the organelle with other parts but is not involved in the gliding itself , because the organelle tends to glide away leaving other parts of the cell [51] . ( iv ) On the surface , P1 adhesin is distributed around the core with P40 and P90 [32 , 52–54] . P30 is localized at the surface of the front end [32 , 36] . This mapping is consistent with the protein clustering based on the fluorescence position and ANOVA test , while some of protein pairs both mapped on the terminal button or the bowl complex were supported for the significant difference in their distribution even in the same localization group ( S3 Table , S6 Fig ) . We suggest a possible model for a gliding mechanism that integrates all information to date ( Fig 6C ) . Movements for gliding may be generated in the bowl complex and transmitted to the paired plates fixed to the cell front through the terminal button . Then , extension and retraction of the organelle repeat . The P1 adhesin complexes connected to the internal core repeat a catch-pull-release cycle with sialylated oligosaccharides on the host surface . The organelle pulls the other parts of the cell connected to the back end of the bowl complex .
The bacterial strains and plasmids are listed in S2 Table . M . pneumoniae strain M129 ( the type strain of subtype 1 ) [55] was used as the wild-type strain and the recipient of transposons . M . pneumoniae cells were grown in Aluotto or PPLO medium at 37°C in a tissue culture flask [7 , 18] . Escherichia coli strains DB3 . 1 , DH5α , and JM83 were used for plasmid construction . For the selection and maintenance of antibiotic-resistant strains of M . pneumoniae and E . coli , antibiotics were used at the following concentrations: gentamicin , 18 μg/ml; chloramphenicol , 15 μg/ml; ampicillin , 50 μg/ml; kanamycin , 50 μg/ml . In a tissue culture flask , M . pneumoniae cells from 0 . 5 L culture in the exponential phase were washed three times with phosphate-buffered saline ( PBS ) consisting of 75 mM sodium phosphate ( pH 7 . 3 ) and 68 mM NaCl . The cells attached to the bottom of the flask were suspended to be a calculated optical density at 600 nm of 20 in PBS containing 0 . 1 mM phenylmethylsulfonyl fluoride ( PMSF ) , and then 1% ( vol/vol ) Triton X-100 and 0 . 1 mg/ml DNase I from bovine pancreas ( Sigma ) were added . After gentle shaking for 5 min at RT , the lysate was put on the top of the sucrose layers , subjected to a stepwise density gradient consisting of 0% , 20% , 30% , 40% , 50% , and 60% sucrose layers in PBS , and centrifuged at 25 , 000 × g for 20 min in a 1 . 5 ml microtube rotor , MX-100 ( TOMY , Tokyo , Japan ) . A fraction of 0 . 2 ml was recovered from the bottom of layer 40% as the core fraction . Using TCA precipitation , the layers of 0% , 20% , and 30% sucrose were combined and the proteins were recovered as a soluble fraction . The whole cell suspension , soluble fraction , and core fraction were subjected to SDS-PAGE and the protein bands were identified by peptide mass fingerprinting ( PMF ) , as described previously [56] . The cell suspension and core fraction were placed directly onto carbon-coated , glow-discharged grids for 5 min at RT . The grids were treated by 0 . 01–1% Triton X-100 or 1% Tween 20 , 1 mg/ml DNase , and 5 mM MgCl2 in PBS for 30 s at RT . After the removal of the solution , the grid was washed with PBS , stained by 2% ammonium molybdate ( vol/vol ) , air-dried , and observed by EM as previously described [6 , 56] . For immuno-gold EM , the culture of M . pneumoniae cells was put onto a grid and incubated for 5 min at RT [57] . After the liquid was removed , the grid was treated with 1% Triton X-100 and 0 . 1 mg/ml DNase in PBS for 1 min . Cells were labeled with 1/100 diluted antiserum against GFP ( Thermo Fisher Scientific , Waltham , MA , USA ) in PBS containing 2% BSA ( PBS-B ) as the primary antibody , and washed five times by PBS . Subsequently , the grids were treated with 1/10 diluted 5 nm colloidal-gold-labeled goat antibody ( Sigma ) in PBS-B for 30 min at RT as the secondary antibody , washed five times by PBS , and then stained negatively with the ammonium molybdate . M . pneumoniae cells expressing EYFP-fused protein were observed under an inverted microscope , IX71 ( Olympus , Tokyo , Japan ) . Images were recorded with a CCD ( charge-coupled device ) camera ORCA-R2 ( Hamamatsu Photonics , Hamamatsu , Japan ) and DP30BW ( Olympus ) . A culture of 2 ml was grown in a glass-bottom dish ( Iwaki , Tokyo , Japan ) and observed directly . Immunofluorescence microscopy was done using the anti-P1 antibodies , as previously reported [19 , 31 , 32] . As a standard of image density , carboxylated polystyrene beads of 200 nm diameter ( Polysciences , Warrington , PA , USA ) were used . The beads were suspended in PBS containing 2 mM MgCl2 at 1/1000 the concentration of the original stock , put on the glass after the cell culture was removed , and kept for 5 min at RT . The cells and beads were washed three times with PBS containing MgCl2 before observation . Image data were acquired by AQUACOSMOS software ( Hamamatsu Photonics ) and analyzed by Image J 1 . 45s ( http://rsb . info . nih . gov/ij/ ) and IGOR Pro ver 6 . 2 ( WaveMetrics , Lake Oswego , OR , USA ) . | Human mycoplasma pneumonia , an epidemic of which occurred around the world a few years ago , is caused by a pathogenic bacterium , Mycoplasma pneumoniae . This tiny bacterium , about 2 μm long , infects humans by gliding on the surface of the trachea through binding to sialylated oligosaccharides , which are also the binding targets of influenza viruses . The mechanism underlying Mycoplasma "gliding motility" is not related to any other well-studied motility systems , such as bacterial flagella and eukaryotic motor proteins . Here , we isolated the internal structure of “attachment organelle" , a cellular architecture , and suggested novel component proteins . The organelle was analyzed systematically by focusing on the protein components under fluorescence and electron microscopy , and a possible gliding mechanism was suggested . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2015 | Systematic Structural Analyses of Attachment Organelle in Mycoplasma pneumoniae |
Understanding how protein structures and functions have diversified is a central goal in molecular evolution . Surveys of very divergent proteins from model organisms , however , are often insufficient to determine the features of ancestral proteins and to reveal the evolutionary events that yielded extant diversity . Here we combine genomic , biochemical , functional , structural , and phylogenetic analyses to reconstruct the early evolution of nuclear receptors ( NRs ) , a diverse superfamily of transcriptional regulators that play key roles in animal development , physiology , and reproduction . By inferring the structure and functions of the ancestral NR , we show—contrary to current belief—that NRs evolved from a ligand-activated ancestral receptor that existed near the base of the Metazoa , with fatty acids as possible ancestral ligands . Evolutionary tinkering with this ancestral structure generated the extraordinary diversity of modern receptors: sensitivity to different ligands evolved because of subtle modifications of the internal cavity , and ligand-independent activation evolved repeatedly because of various mutations that stabilized the active conformation in the absence of ligand . Our findings illustrate how a mechanistic dissection of protein evolution in a phylogenetic context can reveal the deep homology that links apparently “novel” molecular functions to a common ancestral form .
By sequencing genomes of taxa occupying key positions in the metazoan tree of life , it has become possible to infer when important animal gene families originated and proliferated [1]–[3] . Sequence data alone , however , cannot yield insight into the functions and structures of ancient proteins or the processes by which their descendants evolved . Further , many gene families have diversified so extensively that comparisons of extant proteins from model organisms are insufficient to reveal which functions are ancestral and which are derived . In principle , it should be possible to reconstruct the history of a protein family by phylogenetically analyzing the underlying structural mechanisms that produce functional diversity among densely sampled members of the family . Such a strategy would be analogous to detailed studies of the evolution of animal development , which have revealed the deep homology of diverse morphologies in distant lineages and the mechanisms by which they evolved from common ancestral forms [4] . The members of the superfamily of nuclear receptor ( NR ) transcription factors , for example , are regulated in diverse ways—by ligands , postranslational modifications , and association with other proteins or DNA—depending on the cellular context [5] . NRs have a modular domain structure , including a highly conserved DNA-binding domain ( DBD ) and a moderately conserved ligand-binding domain ( LBD ) —which in most receptors contains a ligand-regulated transcriptional activation function—along with extremely variable hinge and N-terminal domains . There is considerable diversity in the functions of NR LBDs , which can be roughly classified into three major groups . In one class , the LBD's transcriptional function can be activated by a specific hydrophobic ligand , such as a steroid , retinoid , or fatty acid; the ligand binds in a deep internal cavity , remodeling and stabilizing the LBD's conformation to generate a new binding surface for coactivator proteins , which increase transcription of nearby genes [5] . The second class of NRs are ligand-independent transcription factors , often called “constitutive” receptors , the LBDs of which can adopt the active conformation and activate gene expression in the absence of a ligand or other modifications . Some members of this class lack the internal cavity and are not known to bind any ligands , whereas others do bind hydrophobic molecules , which up- or down-regulate their baseline activity [6]–[9] . In the third class of NRs , the LBD lacks the capacity to interact with coactivators , so these receptors function primarily as transcriptional repressors that occupy NR response elements or dimerize with and thereby silence other NRs [10]–[12] . It is widely believed that the NR superfamily evolved from a ligand-independent transcriptional activator , with binding of different ligands gained independently in numerous NR lineages [13] , [14] . The alternate view—that NRs evolved from a liganded ancestor , with ligand-dependence lost in the lineages leading to the ligand-independent receptors—has received little attention . These two hypotheses exemplify opposite views on the general nature of molecular evolution and the origin of complex functions . The hypothesis that the ancestral NR was ligand-independent implies that a complex molecular function—allosteric regulation of transcription by binding a ligand—evolved de novo many independent times , requiring evolution to repeatedly create novelty and complexity [15] , [16] . In contrast , the hypothesis of a ligand-activated ancestor implies that evolution produced new functions primarily by subtle tinkering with a conserved ancestral mechanism [4] , [17] , which allowed receptors to accommodate new molecular partners or lose dependence on those partners because of mutations that modified or degraded existing functions . Several limitations have impeded rigorous inference about the ancestral NR's characteristics and the diversification of the superfamily . First , the root of the gene family phylogeny has been ambiguous , leaving unknown the location of the ancestor relative to its descendants . Second , the topology of the NR phylogeny has been uncertain , because of limited sequence sampling and/or use of outdated phylogenetic methods . Third , the functions of NRs in taxa branching near the root of the metazoan phylogeny have not been characterized . Finally , whether distantly related NRs with similar functions share homologous or convergent underlying mechanisms has not been determined . Recently acquired information—including genome sequences from basal metazoans and extensive data on NR structures and functions—along with improved algorithms for phylogenetic analysis of large datasets , now allow these barriers to be overcome . Here we report on biochemical , functional , structural , and phylogenetic analyses of the NR superfamily , which allow us to reconstruct the characteristics of the ancestral nuclear receptor and understand how the functional diversity of NR LBDs evolved .
The root of the NR phylogeny has been unknown because of uncertainty about the relative ages of the various NR family members . NRs appear to be a metazoan innovation , because they are absent from the genomes of choanoflagellates , fungi , plants , and prokaryotes . Until recently , however , all fully sequenced animal genomes have come from protostomes and deuterostomes , both of which contain virtually all the major NR subfamilies [18]; these data indicate only that most NR diversity was already established by the time of the protostome-deuterostome ancestor . To determine the most ancient NR lineages , we identified NRs in the newly sequenced genome of the demosponge Amphimedon queenslandica—a representative of the Porifera , the most anciently branching metazoan phylum based on whole-genome phylogenies [19] . We found that the A . queenslandica genome contains two NRs , which we refer to as AqNR1 and AqNR2 . We amplified transcripts of each by polymerase chain reaction , verified their sequences , and analyzed their developmental expression using in situ hybridization . AqNR2 is expressed ubiquitously , whereas AqNR1 is expressed in a range of cells that contact the external environment and possess apico-basal polarity ( Figure S1 ) . We also identified NRs in genomes from two other recently sequenced early-branching lineages , the placozoan Trichoplax adhaerens and the cnidarian Nematostella vectensis , which contain 4 and 17 NRs , respectively ( see also [20] ) . These results point to very limited NR diversification before the origin of the Eumetazoa and indicate that basal metazoan species have the potential to shed light on early NR evolution . To determine the phylogeny of the NR superfamily , we used model-based phylogenetics to analyze a taxonomically diverse database of 275 NR protein sequences ( Figure 1 , Table S1 ) . The alignment includes the DBDs and LBDs of the complete NR complements in 11 sequenced genomes from eight broadly sampled animal phyla , plus NRs from 30 other species strategically chosen to maximize phylogenetic accuracy and minimize redundant signal [21] . Unlike previous studies , which used sparser sequence sampling and/or less powerful methods [14] , [18] , [22] , phylogenetic analysis of this alignment using maximum likelihood yielded a well-resolved phylogeny ( Figures 1 , S2 ) with strong support for the placement of the basal metazoan sequences and for the relationships among most major NR families . ( A few aspects of the topology , however , had weak support , such as whether the SF-1 class has a monophyletic or paraphyletic relationship to the group containing the steroid hormone receptors . ) We also conducted Bayesian Markov Chain Monte Carlo ( BMCMC ) methods using a slightly smaller 174-sequence dataset , assembled by removing sequences at the ends of very long branches and multiple orthologs within the same phylum . This analysis recovered a nearly-identical phylogeny to the maximum likelihood analysis ( Figures 1 , S3 ) . The phylogeny is unlikely to be an artifact of the presence of rapidly evolving sites or taxa . When the 40% of sites with the fastest evolutionary rates were removed from the analysis , only the placement of the RXR group was affected ( Figure S4 ) . Further , maximum likelihood analysis of the reduced 174-sequence dataset—from which the longest terminal branches were removed—yielded the same phylogenetic relationships as the unreduced analysis ( Figure S5 ) . The phylogeny can be rooted at a single most parsimonious location between AqNR1 and AqNR2 , allowing the ancestral NR to be located on a specific branch of the phylogeny . This rooting provides a coherent history of NR expansion by gene duplication with few subsequent losses ( Figure 2 ) . All alternative rootings that place other NR lineages in a basal position require many additional duplications and losses ( Figure 2A ) . For example , placing the clade of ligand-independent estrogen related receptors ( ERRs ) as the outgroup requires two additional duplications and 12 additional losses compared to the optimal root; placing the ligand-independent NR4 class as the outgroup requires two additional duplications and 15 additional losses . The phylogeny indicates that AqNR2 is orthologous to the fatty acid-binding HNF4 family and that AqNR1 is the unduplicated ortholog of all other NRs . This phylogeny indicates that the last common ancestor of all Metazoa contained two NRs—one ortholog of HNF4 and one of AqNR1 , which subsequently gave rise to all other NR classes ( Figure 2 ) . After the divergence of demosponges from other metazoans but before the split of Cnidaria from the Bilateria , nine more duplications gave rise to most of the major recognized NR types , except for those in the named classes NR1 and NR4 , which proliferated during the interval between the cnidarian-bilaterian ancestor and the protostome-deuterostome ancestor ( Figure 2 ) . Many NR subfamilies diversified further within the vertebrates . Support for the placement of AqNR2 with the HNF4s and of AqNR1 as sister to all other NRs is strong , with posterior probabilities of 1 . 0 and 0 . 98 , chi-square confidence values of 1 . 0 and 1 . 0 , and approximate likelihood ratios of 471 and 112 , respectively ( Figures 1 , S2–S3 ) . The next-best rearrangement of the relationships between the sponge receptors and the rest of the NRs has a likelihood several orders of magnitude lower than that of the ML tree . This alternate phylogeny ( Figure S6 ) would place AqNR1 and AqNR2 as sister paralogs specific to the sponge lineage . It would imply that the ancestral metazoan contained a single NR; duplication of this ancestral gene in the sponges would have yielded AqNR1 and AqNR2 , and the first duplication that separated the HNF4 group from the rest of the NR superfamily would have occurred in the Eumetazoa after they diverged from demosponges . The rest of the superfamily's history would remain unchanged . Hypotheses concerning the functions of the ancestral proteins are strongly affected by the functions of proteins that branch off the family phylogeny near its root . To understand how the functions of the NR LBD evolved , we experimentally characterized the capacity of AqNR1 and AqNR2 LBDs to bind and be regulated by ligands . In a reporter gene assay , AqNR1 had very weak intrinsic activity—less than 2-fold activation—when incubated with serum from which hydrophobic small molecules were stripped using dextran-charcoal . When treated with complete serum , however , AqNR1 transcription increased by 30-fold , suggesting that the receptor is activated by a hydrophobic ligand that is present in mammalian serum , such as a fatty acid or steroid ( Figure 3A ) . To characterize AqNR1's potential ligand , we expressed and purified AqNR1-LBD in bacteria , extracted bound hydrophobic molecules , and used mass spectrometry to identify the bound compounds . Like mammalian HNF4 [23] , AqNR1-LBD bound an array of bacterial free fatty acids ( FAs ) with tail lengths ranging from 16 to 19 carbons , with preference for 18∶0 and 18∶1 fatty acids ( Figures 3B , S7 ) . When AqNR1 was incubated with complete mammalian serum , palmitic acid was the dominant FA bound , along with lower proportions of 18∶0 and 18∶1 FAs ( Figure 3B ) . We then confirmed and quantified FA binding by the purified AqNR1-LBD using an enzymatic assay . As predicted , we found that AqNR1-LBD binds both E . coli and mammalian FAs; receptor occupancy by FAs approximately doubles when the protein is treated with complete serum but does not increase upon treatment with stripped serum ( Figure 3C ) . To more directly test the hypothesis that AqNR1 binds and is activated by FAs , we characterized the functional effects of mutations in the predicted AqNR1 ligand pocket . We first predicted the structure of AqNR1-LBD using a homology model based on the X-ray crystal structure of mammalian HNF4 , the NR with the highest sequence similarity to AqNR1 . The predicted structure ( Figure 4A ) indicates that AqNR1 is likely to have a large ligand pocket ( 835 Å3 ) with ample space to accommodate FAs . As in the crystal structure of mammalian HNF4s , the FA in AqNR1 is predicted to be coordinated by a hydrogen bond from a conserved arginine ( Arg226 in rat HNF4α , Arg492 in AqNR1 ) to the FA's carboxylate oxygen; further , packing interactions between hydrophobic amino acids that line the pocket and the ligand's tail are also conserved . We then used directed mutagenesis and functional assays to experimentally test the hypothesis that AqNR1 binds and is activated by FAs in a manner conserved with mammalian HNF4 . As predicted , when the basic Arg492 was mutated to alanine , reporter activation by AqNR1 in the presence of complete serum was abolished ( Figure 4B ) , and FA binding by the purified protein was dramatically reduced ( Figure 3C ) . Replacement of Arg492 by several other amino acids , each of which would remove the hydrogen bond to the FA's carboxyl oxygen , also abolished reporter activation by AqNR1 ( Figure 4B ) . Although rat HNF4α is a weaker activator in this cell line than AqNR1 , mutations at this site in HNF4α also reduced activation , consistent with a common structural mode of ligand-binding ( Figure 4B ) . Rat HNF4α , which is thought to be activated by fatty acids produced endogenously in liver cells [24] , is not further activated by complete serum , indicating that its specific ligand is different from that of AqNR1 ( Figure 4B ) . We also mutated several hydrophobic residues in AqNR1 that are predicted to contact the FA's tail; as expected , activation by complete serum was dramatically reduced ( Figure 4B ) . One bulky mutation in the predicted pocket , I444W , conferred strong activity on AqNR1 in the presence of stripped or full serum , implying that this mutation stabilizes the active conformation without ligand or allows binding of an unknown ligand that is present in the cultured cells or medium . Taken together , these data indicate that AqNR1's transcriptional activity is affected by binding of a hydrophobic ligand , that the ligand may be a FA , and that key aspects of the AqNR1's structure-function relations are largely conserved with those of mammalian HNF4 . Identification of the specific natural ligand for AqNR1—like that of the ligand for mammalian HNF4 [25]—requires further research , as does determination of whether that ligand has an endogenous or exogenous source . Purified AqNR2 also bound fatty acids ( Figure 3B ) . It did not , however , activate transcription in the mammalian reporter assay but acted as a very strong repressor of basal transcription , irrespective of the type of serum used ( Figure 3A ) . These results indicate that AqNR2 can repress transcription and , like its ortholog HNF4 and its paralog AqNR1 , bind FAs . We cannot rule out the possibility that AqNR2 may have the capacity to activate transcription in the presence of some yet unknown ligand . A robust rooted protein family phylogeny , functional data on basally branching receptors , and recently gathered information on the functions and structures of NRs from model and non-model organisms allow us to infer the characteristics of the ancestral NR . Although the sequences of the NR superfamily are too divergent to allow unambiguous reconstruction of the ancestral NR LBD at the amino acid level , there is substantial phylogenetic signal in the structural and functional features of NR LBDs . We coded these features as discrete phylogenetic characters and reconstructed the best-supported ancestral states using phylogenetic methods ( Figure 5A ) . The ancestral NR ( AncNR ) is decisively reconstructed as having had the capacity to activate transcription , bind a ligand , and be activated by that ligand . The vast majority of extant NRs , including those in the basal lineages , have these characteristics . The handful of exceptions—ligand-independent activators and pure repressors—are scattered across the tree and are in most cases nested deep within groups of liganded activators , indicating that these states are almost certainly derived . The fact that some ligand-independent receptors bind ligands , which can up- or down-regulate their baseline activity [6]–[8] , further supports the reconstruction of the ancestor as having possessed these features . No ligand-independent activators are present in the basally branching NR clades . When the evolution of these functional characters is traced on the NR phylogeny , the hypothesis of a ligand-binding and ligand-activated AncNR is by far the most parsimonious reconstruction . This scenario explains the characteristics of the entire NR superfamily with only five losses of dependence on ligand . Three of these losses were accompanied by a loss of ligand-binding; in the other two instances ( the ERRs and constitutive androstane receptor , CAR ) , receptors evolved “constitutive” transcriptional activity but retained the ancestral capacity to bind ligands , which regulate that baseline activity ( Figure 5A ) . In contrast , the alternative hypothesis of a ligand-independent AncNR would require both ligand-binding and dependence on the ligand for activation to have been gained 12 independent times , plus a subsequent reacquisition of ligand-independent activity in one lineage ( Figure S8 ) . Reconstruction of these characters on the alternate phylogeny that places AqNR1 and AqNR2 as sponge-specific duplicates causes no change in the support for an ancestral liganded-activated receptor vis-à-vis the “constitutive ancestor” hypotheses ( Figure S9 ) . It is also clear that AncNR had the capacity to activate transcription rather than acting as a pure repressor . An ancestral activator requires five losses of activity in the lineages leading to the inactive repressor NRs ( Figure 5A ) , whereas 11 independent gains of transcriptional activity would be required if AncNR were transcriptionally inactive . A key element of assessing homology is to determine whether shared structures and mechanisms underlie apparently similar features in different lineages . To further test the hypothesis that ligand-binding and activation are homologous functions derived from the ancestral NR—and that ligand-independent activation was repeatedly derived—we analyzed the underlying structural mechanisms for these functions in a phylogenetic framework . We coded as discrete phylogenetic characters the relevant structural features of NR LBDs and phylogenetically reconstructed the best supported ancestral state for each ( Figure 5A ) . AncNR is decisively reconstructed as having had the shared features of extant ligand-activated receptors that underlie ligand binding and activation . Specifically , there is strong support for the ancestor having possessed ( 1 ) the classic NR fold in the active conformation consisting of three layers of helices in highly conserved positions; ( 2 ) an open ligand pocket with volume of at least 300 Å3 , bordered by helices H3 , H4-5 , H7 , H10 , and H12; and ( 3 ) a surface for binding coactivator proteins , composed of residues in the ligand-stabilized helices H3 , H4-5 , and H12 , with a conserved coactivator-recognition motif in the latter [5] , [26] , [27] . These features and a similar location of the ligand within a highly conserved LBD structure are shared by even the most distantly related ligand-activated NRs ( Figure 5B ) ; indeed , even the ligand-independent receptors retain some or all of these features . The identical structural basis for ligand-activation throughout the superfamily provides strong evidence that this function is derived from the common NR ancestor . It is plausible that the ancestral ligand was an FA , because several of the most basal lineages bind FAs . Further , the key hydrogen bond between the FA's carboxyl-group oxygen and the Arg side chain on helix 5 is conserved in several basal lineages , including HNF4s , RXRs , and AqNR1 . The ligand that historically activated AncNR could have been a ubiquitous endogenous molecule that served as a receptor cofactor , a hormone-like regulatory compound endogenously produced under specific conditions , or an exogenous nutrient or other substance , such as fatty acids produced by bacteria or other species . In contrast , the structural elements that appear to confer ligand-independence differ dramatically among the ligand-independent activators ( Figure 6 ) . In Nurr1/DHR38 , the mollusk estrogen receptor , and the vertebrate ERRs , the pockets are filled with multiple bulky hydrophobic side chains that mimic the presence of ligand [7] , [8] , [28] , [29] , but the sites and states involved in the three receptor classes are all different , with a single convergent exception in two of the three receptors ( Figure 6A ) . In Drosophila Ftz-F1 , the H6/H7 region adopts an unprecedented loop conformation that turns inward and fills the cavity ( Figure 6A ) . In CAR , residues in helix 12 ( H12 ) form unique hydrogen bonds to H4-5 , and a novel helix , absent from other NRs , packs against H12 , stabilizing the active conformation ( Figure 6B ) [30]–[32] . Finally , in the crystal structure of mouse LRH-1 , the active conformation is stabilized without ligand due to a unique salt bridge between residues in H7 and H10 , which replaces a similar bridge between the ligand and H10 in orthologs of the same protein of humans and other species , and in SF-1 , the closest paralogous NR ( Figure 6C ) [33] . These radical differences in putative underlying mechanisms indicate that ligand-independent activity is a convergent character with independent evolutionary origins rather than a homologous feature inherited from the common NR ancestor . The hypothesis of a ligand-dependent AncNR explains the structure-function relations of the vast majority of present-day receptors as due to descent from an ancestor that possessed those same features . In contrast , the hypothesis of an unliganded AncNR can explain the structure-function relations of only a single NR as due to descent from the ancestral NR; it requires the ancestral basis for ligand-independent activity to have been independently lost and replaced with different underlying mechanisms in all other lineages of ligand-independent receptors and the shared mechanisms for ligand-dependent activation to have been gained independently in the many lineages of liganded receptors . Our findings indicate that NR LBDs evolved their functional diversity by tinkering with a ligand-dependent transcriptional activator . Ligand-regulated NRs are thermodynamically tuned so that in the appropriate contexts the active conformation is favored in the presence of activating ligand but not its absence . The most common functional shift during NR evolution was modification of ligand specificity due to subtle changes in the shape and surface properties of the ancestral ligand pocket . Both historical and contemporary studies indicate that such shifts in ligand preference can evolve through a relatively small number of mutations that subtly alter the ligand cavity ( e . g . , [34]–[36] ) . In a few lineages , ligand-independent activity evolved by mutations that stabilized the active conformation in the absence of ligand; in two such cases , the cavity remained open , yielding a receptor whose baseline activity can be antagonized or super-activated by ligands [6] , [8] . Laboratory and clinical data contain many examples of ligand-independent activity evolving via single point mutations that add sufficient stability to the active conformation in the absence of ligand ( Table S2 ) . Historical studies also document the evolution of constitutive activity with a very simple genetic basis [33] . Such transitions tip the thermodynamic balance so that the formerly switchable LBD becomes stuck in the “on” position , irrespective of ligand . In contrast , evolving a ligand-dependent receptor from a ligand-independent ancestor would require mutations that ( 1 ) generate a ligand pocket of the appropriate size and shape to accommodate some ligand and ( 2 ) destabilize the active conformation just enough to abolish ligand-independent activity but not so much that the capacity is lost to activate transcription when ligand is present . We observed no such transitions on the NR phylogeny , and we are aware of only one NR mutation that accomplishes this end in the laboratory; that example reflects a return to the ancestral amino acid state in a receptor that binds ligand but also possesses ligand-independent activity [37] . Finally , in a few other lineages , inactive repressor NRs evolved by degradation of the activation function without loss of other functions , such as DNA binding , dimerization , or corepressor binding ( see [11] ) . Indeed , most inactive NRs have simply lost the co-activator interaction motif in H12 but retain the classic LBD secondary and tertiary structure , and some even retain an open pocket [12] . Inactive repressor NRs have been shown to evolve from ligand-activated precursors via simple genetic mechanisms that disable ligand or coactivator binding but leave intact other functions of the receptors , such as DNA binding , dimerization , and corepressor binding [11] . Most gene families , like the NRs , have some common conserved core function—some catalytic activity , for example , or the capacity to interact with DNA . Functional diversity within such families is conferred by members' binding to and carrying out that function on different partners . Our observations in the NRs underscore the capacity of evolution to produce dramatic functional diversity by tinkering with a common ancestral template over long periods of time . The varied and subtle nature of these tinkering events is revealed only when densely sampled structural and functional data are analyzed in a phylogenetic context . We predict that , when sufficient data are gathered to allow detailed evolutionary reconstructions , it will become apparent that most protein superfamilies diversified by subtle modification and partial degradation of ancient , deeply homologous functions . Invoking the evolution of wholesale “novelty” will seldom be necessary .
Nuclear receptor protein sequences were obtained by mining the genomes of Amphimedon queenslandica , Trichoplax adhaerens , Nematostella vectensis , Lottia gigantea , Capitella capitata , and Branchiostoma floridae ( Table S1 ) . The assembled genomes and developmental expressed sequence tags were screened using tBlastn with LBD and DBD amino acid sequences from each known NR family . Further analysis using PFAM domain analysis ( PF00104 and PF00105 ) [38] and a hidden Markov model-based method ( PTHR11865 ) [39] confirmed the presence of only two NRs in the A . queenslandica genome , which has been sequenced at approximately 9-fold coverage [19] . In some of the other genomes , gene model sequences were modified to resolve gaps in the sequence by performing a local assembly with gene traces or to correct the predicted protein sequence based on alignment with other conserved domain sequences . Complete NR complements from the curated whole-genome databases of H . sapiens , D . melanogaster , C . intestinalis , F . rubripes , and S . purpuratus were also included . Additional nuclear receptors were identified by using the SMART domain-based sequence annotation resource [40] to search the UniPROTKB/TrEMBL database based on the amino acid sequence of the ERR of Marisa cornuarietis . Receptors for which only partial sequence was available ( missing >20% of the DBD or LBD ) and those entirely lacking a DBD domain ( e . g . , human DAX1 and SHP ) or LBD domain ( e . g . , D . melanogaster Knirps and Knrl ) were excluded from the analysis . A total of 275 nuclear receptor sequences were aligned . Full-length sequences containing the DBD , highly variable hinge region , and LBD were aligned using Multiple Sequence Alignment by Log-Expectation ( MUSCLE ) v . 3 . 6 [41] in order to identify the boundaries of the conserved regions . After removal of the variable ( non-alignable ) hinge region , sequence blocks corresponding to the DBD and LBD were then aligned separately using MUSCLE . The DBD and LBD alignments were checked manually to remove lineage-specific indels , and the LBD alignment was checked to ensure correct alignment of the conserved AF-2 core sequence ( φφ*κφφ motif; φ , hydrophobic; * , any residue; κ , charged ) . Amino acids C-terminal to this AF-2 core sequence could not be reliably aligned among all nuclear receptors and the LBD alignment was therefore truncated after the AF-2 core sequence . The DBD and LBD alignments were then concatenated in MacClade 4 ( Sinauer Associates , Inc . , MA , USA ) for subsequent phylogenetic analyses . We then used APDB software [42] to characterize the quality of our alignment with reference to the 26 NR LBDs in the alignment for which X-ray crystallographic structures are available; the average iRMSD ( the root mean square difference of the intramolecular distances between aligned pairs of alpha-carbons ) over the entire LBD alignment was 0 . 82 angstroms , well under the resolution of the structures themselves , indicating that the alignment has high structural plausibility . Phylogenetic analyses were performed using maximum likelihood in PhyML v . 2 . 4 . 5 [43] and Bayesian analysis using MrBayes v . 3 . 1 [44] . The Jones-Taylor-Thornton model with a four-category discrete gamma distribution of among-site rate variation ( ASRV ) and a proportion of invariant sites was used . For ML , all model parameters were optimized by maximum likelihood . Support was evaluated by obtaining the approximate likelihood ratio for each node—the estimated ratio of the likelihood of the best tree with the split to the best tree without the split [45]—as well as the chi-square confidence metric , which approximates 1−p , where p is the probability that an approximate likelihood ratio as great or greater than that observed at a resolved node would occur if the null hypothesis of an unresolved node is true [45] . To identify the next best alternative tree for the basal split between the AqNR1 and AqNR2-containing groups , we used Phyml to optimize the branch lengths and model parameters on each of the two possible rearrangements of the ML tree around this internal branch and then report their likelihoods . To determine the effect of fast-evolving sites on the inference of phylogeny , we used PAML software to identify sites in the top two octiles of the gamma distribution ( Table S3 ) and repeated the analysis with those 113 sites removed . To facilitate adequate sampling of tree space in Bayesian analysis [46] , we used a 174-sequence taxon-trimmed MUSCLE-aligned dataset including nuclear receptors from the following taxa representative of the major metazoan lineages: Acropora millepora , Nematostella vectensis , and Tripedalia cystophora ( cnidarians ) ; Amphimedon queenslandica and Suberites domuncula ( poriferans ) ; Branchiostoma floridae ( cephalochordate ) ; Capitella capitata and Lottia gigantea ( lophotrochozoans ) ; Ciona intestinalis ( urochordate ) ; Drosophila melanogaster ( ecdysozoan ) ; Homo sapiens ( vertebrate ) ; Saccoglossus kowalevskii ( hemichordate ) ; Strongylocentrotus purpuratus ( echindoderm ) ; and Trichoplax adhaerens ( placozoan ) . Terminal branches of length ≥0 . 76 in the PhyML analysis ( except for AqNR2 ) were removed . Four heated chains were run for 8 million generations with temperature 0 . 3; the cold chain was sampled every 100 generations . Priors were uniform on topologies , uniform ( 0 , 5 ) on branch lengths , and uniform ( 0 . 1 , 10 ) on the alpha shape parameter . The first 6 , 694 , 000 generations were discarded as burn-in , because at this point in the chain the standard deviation of posterior probabilities over all splits was <0 . 01 and the two chains had converged as evaluated using the “compare” option of AWTY software [47] . We also repeated ML analysis on this reduced 174-sequence dataset and found no change in the relationships among NR types ( Figure S5 ) . The phylogeny shows that AqNR1 is the ortholog of the previously identified but misnamed “RXR” gene identified in the sponge Suberites domuncula [48] and is identical to the “HNF4” gene previously reported in A . queenslandica [49] . To determine the minimum number of gene duplications and losses , we used SDI software [50] . The 275-sequence ML phylogeny was reduced by collapsing sets of orthologous NRs within major taxa—Porifera , Placozoa , Cnidaria , Protostomia , Deuterostomia—into single clades . To avoid spurious inference of duplication/loss due to phylogenetic error , nodes with likelihood-ratio support <10 that conflicted with the accepted taxonomic phylogeny ( Porifera , ( Placozoa , ( Cnidaria , ( Protostomia , Deuterostomia ) ) ) ) were treated as unresolved and reordered to be congruent with the taxonomic phylogeny . SDI software was then used to reconcile this gene family tree with the taxonomic tree and identify the root with the lowest possible mapping cost ( duplications plus losses ) . The mapping cost was also calculated for all possible roots of the gene family phylogeny , except for rootings on branches after the Cnidaria/Bilateria divergence , which have higher mapping costs and were judged to be implausible . Reconstructions of ancestral structural and functional states were performed manually using Fitch parsimony . Demosponge Amphimedon queenslandica were collected from Heron Island Reef , Great Barrier Reef , and total RNA was isolated from larvae using RNeasy Mini kit ( Qiagen , Valencia , CA ) . The coding regions of AqNR1 and AqNR2 were obtained using BD SMART RACE cDNA Amplification kit ( Clontech , Mountain View , CA ) , and the full reading frames were amplified by RT-PCR , cloned into pGEM-T EASY vector ( Promega , Madison , WI ) , and verified by sequencing . In situ hybridization analysis of RNA expression was conducted as previously described [49] . AqNR1 , AqNR2 , and rat HNF4α receptor LBDs , including the hinge region and carboxy-terminal extension , were amplified by high-fidelity PCR using Phusion DNA polymerase ( New England Biolabs , Ipswich , MA ) and cloned into a GAL4-DBD-pSG5 expression vector ( gift of D . Furlow ) . AqNR1-LBD ( gi ACA04755 ) consisted of amino acids 263636 , and AqNR2-LBD ( GU811658 ) included amino acids 118–852 . Rat HNF-4α template was a gift from Frances Sladek; the LBD used consisted of amino acids 116–465 ( NP_071516 ) . Site-directed mutagenesis was performed using QuickChange II ( Stratagene , La Jolla , CA ) and verified by sequencing . Chinese hamster ovary ( CHO-K1 ) cells were grown in a 96-well plate and transfected with1 ng of receptor plasmid , 100 ng of a UAS-driven firefly luciferase reporter ( pFRluc ) , and 0 . 1 ng of the constitutive phRLtk Renilla luciferase reporter plasmid , using Lipofectamine and Plus Reagent in OPTIMEM ( Invitrogen , Carlsbad , CA ) . After 4 h , transfection medium was replaced with phenol-red-free αMEM supplemented with 10% dextran-charcoal-stripped fetal bovine serum ( Hyclone , Logan , UT ) . Cells were allowed to recover and express protein for 48 h , and then assayed by luminometry using the Dual-Glo luciferase system ( Promega , Madison , WI ) . Firefly luciferase activity was normalized by Renilla luciferase activity . AqNR1 LBD ( residues 415–636 ) , AqNR1 mutant proteins , AqNR2 LBD ( residues 616–852 ) , and rHNF4α ( residues 133–382 ) were expressed as N-terminal hexahistidine maltose binding protein fusions with a TEV cleavable linker in pLIC-MBP ( a gift from J . Sondek ) and grown in E . coli BL21 ( DE3 ) pLysS cells using standard methods . Protein was purified using affinity chromatography using standard methods . Following TEV cleavage , the resulting 6xHis-tagged MBP was removed using an additional nickel affinity column and the AqNR1 or AqNR2 was polished via gel filtration . Pure AqNR1 or AqNR2 was dialyzed against 150 mM ammonium acetate ( pH 7 . 4 ) prior to lipid extraction . Organic solvent extraction was performed on purified LBDs from bacteria to facilitate detailed characterization of bound ligands in the absence of protein . Before extraction , 0 . 1 mg of C13 labeled palimitic acid was added as an internal standard . Lipid from approximately 4 mg of wild-type or mutant forms of AqNR1 LBD , AqNR2 , or rHNF-4 LBD were extracted with a 2∶1 chloroform/methanol ( v/v ) solution and then analyzed by negative ion ESI/MS . All extractions were performed in duplicate . Mass spectra were acquired on a LTQ FT Hybrid Mass spectrometer ( Thermo Finnigan LTQ-FTMS , Somerset , NJ ) equipped with an electrospray source . Typically , 10 µL of the aforementioned lipid solution was diluted into 10 µL water/acetonitrile ( 2∶1 v/v ) and subjected to ESI/MS in the negative ion mode . In addition to the fatty acids shown in Figure 2B , an additional unidentified substance at ∼421 m/z was also bound when AqNR1 was incubated with either complete or stripped serum . All samples were run in triplicate . Data acquisition and analysis were performed using the instrument's xcalibur software . Purified wild-type or mutant hexahistidine maltose binding protein fused AqNR1 was incubated with undiluted complete ( Invitrogen - 26010 ) or cyclodextran/charcoal stripped ( HyClone -SH30068 . 03 , Waltham , MA ) serum at a ratio of 20 mg protein to 5 ml undiluted serum . The protein/serum mixture was incubated overnight at 4°C followed by re-purification over a nickel affinity column . Protein purity was assessed by SDS-PAGE and fractions containing pure wt or mutant AqNR1 were pooled . Bound lipids were then quantified using the free fatty acid quantification kit from BioVison Inc . ( Mountain View , CA ) . 0 . 5 mg of each purified LBD was subject to chloroform/detergent extraction to isolate the long chain free fatty acids . Extracted fatty acids were enzymatically converted to their CoA derivatives and oxidized , allowing quantitation in a colorimetric assay ( λ = 570 ) relative to a standard curve generated using palmitic acid . Efforts to determine the crystal structure of AqNR1-LBD were unsuccessful , so its structure was predicted by homology modeling and energy minimization . The AqNR1 LBD amino acid sequence was aligned to and threaded on human HNF-4α ( PDB 1M7W ) and then energy minimized with palmitic acid—the most abundant experimentally bound ligand—using the Homology module in InsightII ( Accelrys , Inc . , San Diego , CA ) . To calculate ligand pocket volumes of receptors with X-ray crystal structures , we used VOIDOO [51] in probe-occupied mode . We assigned the centroid of the bound ligand or a manually defined point as a starting locus for cavity searches . Cavity volumes were calculated using 10 random orientations of the protein using 10 different “van der Waals growth factors” ranging from 1 . 1–1 . 3 . Mean and mode cavity volumes with standard deviation are listed in Table S4 . To calculate ligand pocket volumes of receptors whose structures have not yet been determined by x-ray crystallography , we inferred homology models . Specifically , we created homology models of the LBDs of AqNR1 ( gi 167859601 , residues 404–534 ) , annelid ER ( 186908731 , residues 231–479 ) , Branchiostoma SR ( 170178459 , residues 298–532 ) , and Branchiostoma ER ( 170178461 , residues 250–504 ) . In each case , we used as templates crystal structures of several NR LBDs with a variety of cavity volumes , including human ERα with estradiol ( PDB 1ERE:A , cavity volume 447 Å3 ) , human ERR3 apo form ( 1KV6:A , cavity volume 262 Å3 ) , human ERR1 apo form ( 3D24:A , cavity volume 42 Å3 ) , and AqNR1 as modeled on template HNF4A with DAO ( 1MV7:A , cavity volume 680 Å3 ) . We generated 10 models for every protein with Modeller 9 . 7 using the default parameters [52] . Models were visually inspected for artifacts ( e . g . , knotting ) and further assessed using RamPage in CCP4i software; only models with 95% of residues in the preferred region and <1% of residues in the outlier region of the Ramachandran map were accepted . We then used Voidoo software as described above to calculate cavity volumes , which are listed in Table S5 . | Many protein families are so diverse that it is hard to determine their ancestral functions and to understand how their derived functions evolved . The existence of so many different functions within protein families often creates the impression that complex , novel functions must have evolved repeatedly and independently . Nuclear receptors ( NRs ) are a large family of related proteins that regulate key biological processes in animals by binding to specific DNA sequences and triggering expression of nearby target genes . Many NRs are activated by a specific hormone or other small molecule , but some do not require a ligand , and still others are incapable of activating gene expression and so act primarily as repressors of transcription . To understand how the functional diversity of NRs evolved , we reconstructed the structural and functional characteristics of the ancient protein from which the entire family evolved , using genomic , biochemical , functional , and structural analyses in a phylogenetic framework . We show , contrary to current belief , that the ancestral NR was a ligand-activated transcriptional activator that existed in the earliest period of animal evolution . Our analysis reveals how the extraordinary functional diversity of modern receptors was generated by subtle tinkering with this ancestral template—slightly reshaping the ligand cavity , stabilizing the protein's active conformation so it no longer required a ligand , or disabling the protein's capacity to activate transcription without affecting its other properties . We predict that , when sufficient data are gathered to allow detailed evolutionary reconstructions in other protein families , it will become apparent that most protein functional diversity evolved by tinkering with ancient functions; invoking the evolution of wholesale “novelty” will seldom be necessary . | [
"Abstract",
"Introduction",
"Results/Discussion",
"Methods"
] | [
"evolutionary",
"biology",
"biochemistry/molecular",
"evolution",
"diabetes",
"and",
"endocrinology/reproductive",
"endocrinology"
] | 2010 | Protein Evolution by Molecular Tinkering: Diversification of the Nuclear Receptor Superfamily from a Ligand-Dependent Ancestor |
Protein quality control is essential for clearing misfolded and aggregated proteins from the cell , and its failure is associated with many neurodegenerative disorders . Here , we identify two genes , ufd-2 and spr-5 , that when inactivated , synergistically and robustly suppress neurotoxicity associated with misfolded proteins in Caenorhabditis elegans . Loss of human orthologs ubiquitination factor E4 B ( UBE4B ) and lysine-specific demethylase 1 ( LSD1 ) , respectively encoding a ubiquitin ligase and a lysine-specific demethylase , promotes the clearance of misfolded proteins in mammalian cells by activating both proteasomal and autophagic degradation machineries . An unbiased search in this pathway reveals a downstream effector as the transcription factor p53 , a shared substrate of UBE4B and LSD1 that functions as a key regulator of protein quality control to protect against proteotoxicity . These studies identify a new protein quality control pathway via regulation of transcription factors and point to the augmentation of protein quality control as a wide-spectrum antiproteotoxicity strategy .
Living organisms endure environmental stress and metabolic errors that inflict damage on macromolecules , including DNA and protein , which are either repaired or removed by quality control programs in the cell . Toxicity resulting from protein misfolding and aggregation , known as proteotoxicity , underlies many degenerative diseases , including those affecting the nervous system , such as Creutzfeldt-Jakob disease , Alzheimer disease , Parkinson disease , Huntington disease , frontotemporal dementia ( FTD ) , and amyotrophic lateral sclerosis ( ALS ) [1 , 2] . To guard against proteotoxicity , the cell coordinates several major quality control systems , including molecular chaperones , the ubiquitin-proteasome system , and autophagy [3–6] . The regulation of these systems occurs on different scales , from individual proteins to the whole organism [7] . It is tantalizing to envisage enhancing the protein quality control systems to defend against proteotoxicity associated with neurodegenerative diseases . However , how the protein quality control might be harnessed in the cell to alleviate proteotoxicity-associated neurodegeneration is not yet fully understood . Mutant Cu/Zn superoxide dismutase ( SOD1 ) , linked to ~20% of familial ALS , represents a simple molecular model for protein misfolding and aggregation . The wild-type ( WT ) SOD1 protein has a stable β-barrel structure with a two-state folding process , whereas mutant SOD1 proteins show a heightened propensity to aggregate in vitro and in vivo [8–10] . There is increasing evidence that heightened propensity to misfold and aggregate is a common feature of ALS/FTD-associated proteins , including TAR DNA binding protein 43 ( TDP-43 ) and fused in sarcoma ( FUS ) [11–13] . Identifying mechanisms that suppress the toxicity of protein misfolding and aggregation may help elucidate the pathogenesis of neurodegenerative diseases and provide potential targets for correction . Here we took advantage of a Caenorhabditis elegans model that expresses neuronal ALS-linked SOD1 mutant proteins and develops robust movement defects to perform an unbiased genetic screen for potent suppressors of the behavioral defects . We identified mutations in two genes , ufd-2 , encoding a ubiquitin ligase , and spr-5 , encoding a lysine-specific demethylase , that synergistically attenuate the neurotoxicity of mutant human SOD1 proteins . The actions of the suppressor genes are conserved in Drosophila , and they protect against proteotoxicity initiated by diverse mutant proteins , including TDP-43 , FUS , and the polyglutamine ( polyQ ) tract . Furthermore , we found human orthologs of these modifiers to be part of a pathway regulating protein quality control in mammalian cells . Further analysis showed that this pathway acts through the transcription factor p53 , which mediates cellular stress responses . Together , these results identify a new mechanism involving previously unrecognized players , which the cell utilizes to augment protein quality control .
To better understand the regulatory mechanisms that mitigate an increased load of misfolded proteins , we conducted a forward genetic screen for cellular factors that alleviate such stress and relieve cells from proteotoxic insults . This screen took advantage of a C . elegans model of ALS , in which the neuron-directed expression of the ALS-linked , G85R mutant human SOD1 ( SOD1G85R ) protein leads to its aggregation into misfolded soluble oligomers and larger insoluble aggregates [14 , 15] . Misfolded SOD1G85R protein is highly toxic , leading to age-dependent synaptic dysfunction , neurodegeneration , and severely impaired movement in the worms [14] . This severe locomotor defect allowed us to perform a large-scale screen for genes that suppress neurodegeneration and improve worm locomotion . In these experiments , we treated homozygous transgenic SOD1G85R C . elegans with ethyl methanesulfonate ( EMS ) to induce genomic mutations , and the mutagenized P0 hermaphrodites were allowed to self-reproduce for two generations ( Fig . 1A ) . Next , in the F2 offspring , which contain both heterozygous and homozygous suppressor mutations , we selected individual C . elegans based on a salient improvement in the locomotion on a background of poorly moving populations . The potential suppressor clones were bred through until 100% of progeny showed phenotypic improvements and were then subjected to further analysis ( Fig . 1A ) . After screening >105 haploid genomes , we isolated hundreds of independent strains with markedly improved locomotion . Most of these strains were dismissed upon closer examination because they showed a reduction in the expression of a green fluorescent protein ( GFP ) reporter gene that had been coinjected as an internal reference and expressed independently in the pharynx , suggesting silencing of the transgene cassette . Among the few remaining suppressor strains that survived this test , one designated M1 showed potent suppression of the locomotion defect when compared with the parental SOD1G85R line , reaching ~76% of the locomotion robustness of the SOD1-WT transgenic line ( Fig . 1B and S1 Movie ) . Such strong recovery of locomotion was apparently not a consequence of diminished SOD1G85R transgene expression because SOD1G85R mRNA and protein levels were unchanged between the parental and M1 mutant strains ( Fig . 1C ) . Further segregation analysis of M1 indicated that more than one genetic locus , in addition to the SOD1 transgene on chromosome IV , was linked to the suppressor phenotype , suggesting a rare multigenic suppressor underlying the suppressor phenotype . To map and identify genes responsible for the suppression of the locomotor defect , we carried out single-nucleotide polymorphism ( SNP ) mapping [16] . SNP mapping localized the M1 suppressor mutations to two linkage regions: a 2 . 2-Mb interval on chromosome I and an 8-Mb interval on chromosome II ( Fig . 1D ) . Next , we performed two rounds of deep sequencing on the M1 strain genomic DNA [17] , attaining a 27-fold coverage . When the M1 genomic DNA sequencing data was aligned with the C . elegans reference genome , we found over 200 variants in the two linkage regions . Next , we performed deep sequencing of the parental strain carrying only the SOD1G85R transgene , with 7 . 5-fold coverage . Comparison of the parental and M1 genomic sequences indicated that most of the nonreference variants existed prior to the EMS mutagenesis and thus were not responsible for the suppressor phenotype . Our analysis pinpointed two variants as likely candidates for the suppressor mutations in M1: in the chromosome I linkage region , there is only one missense mutation , G1937A , resulting in a single amino acid change ( R646Q ) in the gene suppressor of presenilin 5 ( spr-5 ) ; and on chromosome II , among the few remaining variants is one nonsense mutation , G2472A , which results in a premature stop ( W824X ) in the gene ubiquitin fusion degradation 2 ( ufd-2 ) ( Fig . 1D and 1E ) . To examine the role of the double mutations ufd-2 ( W824X ) and spr-5 ( R646Q ) in the suppression of mutant SOD1-mediated neurotoxicity , we performed a series of genetic , biochemical , and behavioral analyses . ufd-2 encodes a U-box type ubiquitin ligase , and the W824X mutation results in a truncated protein lacking the C-terminal U-box ( Fig . 1E ) . spr-5 encodes a lysine-specific demethylase , and the R646Q substitution occurs at a highly conserved residue in the C-terminal portion of an amine oxidase-like ( AOL ) domain ( Fig . 1E ) . While either ufd-2 ( W824X ) or spr-5 ( R646Q ) alone did not lead to the strong locomotor defect-suppressing phenotype in the M1 strain , the double mutation ufd-2 ( W824X ) and spr-5 ( R646Q ) segregated perfectly with the M1 phenotype , recapitulating the full rescuing effect of the suppressor . To confirm ufd-2 and spr-5 as the suppressor genes , we obtained independent null alleles of the two genes: a deletion mutation , ufd-2 ( tm1380 ) , that lacks 80% of the protein at the C-terminus [18] and a nonsense mutation , spr-5 ( by134 ) , that lacks the C-terminal half of the protein ( Fig . 1E ) [19] . When crossed to the mutant SOD1 strain , the single allele of ufd-2 ( tm1380 ) provided a moderate , 2-fold locomotor improvement , and less improvement was seen for the single allele of spr-5 ( by134 ) ( Fig . 1F ) . However , combining the alleles of spr-5 ( by134 ) and ufd-2 ( tm1380 ) completely recapitulated the strong locomotor-defect-suppressing phenotype observed in the M1 strain ( Fig . 1F ) . Total levels of SOD1G85R protein were similar among the WT , single- , and double-mutant strains ( S1B Fig . ) . However , further analysis after fractionation by solubility revealed that the insoluble level of SOD1G85R , which accounts for less than 2% of total proteins , was decreased by the spr-5 ( by134 ) ;ufd-2 ( tm1380 ) mutations , while the soluble level of SOD1G85R remained unchanged ( Fig . 1G ) . Finally , we found that restoring the function of either ufd-2 or spr-5 alone by expressing transgenic wild-type ufd-2 or spr-5 under a neuron-specific promoter from the synaptobrevin ( snb-1 ) gene completely blocked the protection in the M1 strain ( Fig . 1H ) , indicating that it was the loss of function in these two genes and not any other background mutation that was responsible for the suppressor phenotype . Taken together , these results establish that the synergistic loss of ufd-2 and spr-5 creates a potent novel suppressor of the neurodegenerative phenotypes in the SOD1 C . elegans model of ALS , which we have termed the spr-5– and ufd-2–dependent neurodegeneration suppressor ( SUNS ) . Next , we asked whether the SUNS suppressor genes affect the aggregation and toxicity of other misfolded proteins . In transparent C . elegans models , yellow fluorescent protein ( YFP ) fusions of several disease-relevant , aggregation-prone proteins , such as SOD1G85R-YFP [14] , TDP-43c25-YFP [20] , and PolyQ-YFP [21] , have been used to facilitate the visualization of their protein aggregation . These fusion proteins form fluorescent protein aggregates readily visible in live C . elegans , and , when present in neurons , these protein aggregates correlate with the toxicity to the animals as manifested in their locomotor defects [14 , 20 , 21] . To determine whether the SUNS mutant reduces the toxicity associated with these protein aggregates , we introduced the double-null mutations ufd-2 ( tm1380 ) ;spr-5 ( by134 ) into the strains that pan-neuronally express SOD1G85R-YFP , TDP-43c25-YFP , or PolyQ-YFP . Indeed , loss of ufd-2 and spr-5 function resulted in a marked reduction in the neuronal protein aggregation when compared with controls . Reduction in the number and intensity of protein aggregates was evident in the change in the fluorescent inclusions in the head and ventral cord regions of the SUNS-mutant C . elegans ( Fig . 2A–2C ) . Consistently , the locomotor phenotypes in these C . elegans strains were significantly improved by the introduction of the SUNS mutations , ufd-2 ( tm1380 ) ;spr-5 ( by134 ) ( Fig . 2A–2C ) . To investigate the biochemical states of the misfolded proteins in the C . elegans models , we performed a protein solubility assay by differentially extracting and sedimenting worm lysates into soluble supernatants and insoluble pellets . The worm pellet fraction is enriched with sedimentable large SOD1 protein aggregates , whereas the supernatant fraction contains smaller aggregates and oligomeric species ( S1A Fig . ) [14 , 15] . Western blot analysis of both supernatant and pellet fractions displayed a significant decrease in the levels of misfolded SOD1G85R-YFP , TDP-43c25-YFP , and PolyQ-YFP in the ufd-2 ( tm1380 ) ;spr-5 ( by134 ) double mutant when compared with controls ( Fig . 2A–2C ) . Compared with the untagged SOD1G85R ( the soluble level of which was unchanged , S1B Fig . ) , SOD1G85R-YFP was significantly reduced in its soluble fraction by ufd-2 ( tm1380 ) ;spr-5 ( by134 ) . This suggests that soluble SOD1G85R-YFP proteins were degraded more rapidly than the untagged SOD1G85R , consistent with the notion that the YFP tag may increase the misfolding of fusion proteins and therefore decrease their total protein levels ( S1C Fig . ) . Taken together , these data indicate that the ufd-2;spr-5 double mutations are capable of reducing the proteotoxicity of various misfolded proteins associated with neurodegeneration , suggesting a wide-spectrum effect of the suppressor . To investigate whether this effect was evolutionarily conserved , we assessed the actions of ufd-2 and spr-5 in transgenic Drosophila models expressing known ALS disease-causing human TDP-43M337V or FUSR521C proteins . These models have been previously shown to cause photoreceptor degeneration and rough-eye phenotypes [22 , 23] . We found that knockdown of CG9934 , a Drosophila homolog of C . elegans ufd-2 and human ubiquitination factor E4 B ( UBE4B ) , rescues the TDP-43M337V-induced degenerative rough-eye phenotype , resulting in a smoother eye appearance ( Fig . 2D , left ) and restoration of pigmentation ( Fig . 2E ) . Similarly , knockdown of the ufd-2 homolog CG9934 corrected the ommatidial defects ( Fig . 2D , middle ) and pigmentation loss induced by FUSR521C ( Fig . 2E ) . Additionally , we observed that knockdown of Drosophila Su ( Var ) 3-3 , the homolog of C . elegans spr-5 and human lysine-specific demethylase 1 ( LSD1 ) , rescued the degenerating eye phenotypes induced by FUSR521C ( Fig . 2D , right , and Fig . 2E ) . Neither of the suppressors changed the protein expression levels of TDP-43M337V or FUSR521C ( S1D Fig . ) . Taken together , these results indicate that the loss of ufd-2 and spr-5 homologs also suppresses proteotoxicity-related phenotypes in diverse Drosophila models . Homologs of ufd-2 and spr-5 are present in all eukaryotes . UBE4B and LSD1 are the human orthologs of ufd-2 and spr-5 , respectively . UBE4B and LSD1 share 32% and 29% sequence identity with ufd-2 and spr-5 , respectively , and all the major protein domains are conserved ( Fig . 1E ) . To determine whether UBE4B and LSD1 affect protein aggregation in mammals , we used a protein solubility assay that was established to characterize the aggregation of mutant SOD1 in HEK293T cells ( S2A Fig . ) [9 , 24] . This assay utilizes the aggregation-prone SOD1G85R protein , which migrates faster on SDS-PAGE than its WT counterpart , as a reporter of protein aggregation , with WT SOD1 serving as an internal control . We knocked down UBE4B and/or LSD1 with multiple RNA interference ( RNAi ) oligonucleotides and analyzed the levels and solubility of SOD1G85R protein in HEK293T cells . Cell lysates were subjected to ultracentrifugation to separate insoluble pellets , which contain large and sedimentable SOD1G85R aggregates , from soluble supernatants , which contain correctly folded native proteins , misfolded proteins , and small oligomeric aggregates ( S1A Fig . ) . The WT SOD1 protein remained in the supernatant ( S ) in all tested conditions , but a significant portion of SOD1G85R protein ( 25%–30% ) was enriched in the insoluble pellet ( P ) fraction ( Fig . 3A ) . Knockdown of both UBE4B and LSD1 significantly decreased levels of SOD1G85R in both supernatant and pellet fractions ( Fig . 3B ) , consistent with reduction of total SOD1G85R proteins ( S2B Fig . and S2C Fig . ) . The UBE4B/LSD1 knockdown did not affect the protein levels of WT SOD1 . These results were consistent with the notion that the mutant SOD1G85R had a much larger fraction of misfolded and aggregated proteins that were sensitive to UBE4B- and LSD1-dependent clearance than the WT SOD1 protein . In agreement with the observations in C . elegans and Drosophila , the effects of UBE4B/LSD1 knockdown on enhanced protein clearance was not specific to mutant SOD1G85R but also occurred with other aggregation-prone proteins , including TDP-43Q331K ( S2D Fig . and S2E Fig . ) , indicative of a general effect on protein quality control . Additionally , single knockdown of either LSD1 or UBE4B also resulted in decreased steady-state levels of SOD1G85R , with the UBE4B knockdown producing a stronger effect than LSD1 ( Fig . 3B ) . However , the double knockdown produced an even more pronounced decrease in the SOD1G85R levels , indicating an additive effect between UBE4B and LSD1 . To determine whether this decrease in the levels of misfolded and aggregated proteins was a consequence of increased degradation of proteins , we performed cycloheximide chase experiments . Cells were transfected with SOD1G85R together with either nontargeting control small hairpin RNAs ( shRNAs ) or a mix of UBE4B and LSD1 shRNAs , and the clearance of the SOD1G85R protein was quantified . Cycloheximide was used to block de novo translation , and the amount of SOD1G85R protein remaining in the supernatant at the indicated time points after the translation block was determined by SDS-PAGE and western blotting ( Fig . 3C ) . The UBE4B and LSD1 double knockdown decreased the half-life of SOD1G85R from 8 . 5 h to 5 h , indicating that increased clearance of the mutant protein underlies the reduction of the protein aggregates ( Fig . 3D and 3E ) . Similarly , the reduction of TDP-43Q331K by knockdown of UBE4B and LSD1 was also attributable to increased protein clearance , as shown by cycloheximide chase experiments ( S2F Fig . and S2G Fig . ) , further suggesting that a general enhancement of protein quality control is the consequence of the loss of function of the suppressor genes . To identify the downstream effectors of UBE4B and LSD1 in the antiproteotoxicity pathway , we performed a comprehensive transcriptional analysis using the cell-based SOD1 misfolding model . We treated HEK293T cells with shRNAs targeting either UBE4B or LSD1 alone or UBE4B and LSD1 simultaneously in the presence of SOD1G85R proteins . Upon confirmation of reduction in UBE4B and LSD1 protein levels , total RNA was isolated and subjected to microarray profiling of the whole human transcriptome ( S3A Fig . ) . In triplicate samples of the three knockdown conditions and nontargeting controls , differentially regulated genes and pathways were analyzed in unbiased approaches to identify those that convey the UBE4B/LSD1-mediated activation of protein quality control systems . The most intriguing observation in our unbiased microarray analysis did not concern individually regulated genes but instead related to the upstream regulators that elicited characteristic pattern of mRNA level changes in a whole pathway or network . By employing the Ingenuity Pathway Analysis ( IPA ) algorithm to compare the predicted pattern of changes and the actual changes in these genes in our microarray profiles , we identified a number of upstream regulators whose downstream targets are significantly changed ( z-score ≥ 2 . 0 ) in UBE4B and LSD1 single or double knockdowns ( Fig . 4A ) . Among these upstream regulators , only a few were shared by more than one experimental condition , and remarkably , p53 was the only upstream regulator common to all three conditions ( Fig . 4A and S1 Table ) . In the UBE4B and LSD1 double-knockdown condition , a large number of p53 target genes were affected , and importantly , a large fraction were changed in the directions that statistically suggest an activation of the p53 transcription factor ( Fig . 4B , S3B Fig . , S3C Fig . , S2 Table ) . We examined a sample of 11 p53 target genes in HEK293T cells and confirmed by reverse transcription-quantitative polymerase chain reaction ( RT-qPCR ) that their expression levels were consistent with the microarray dataset ( Fig . 4C ) . Since UBE4B is a ubiquitin ligase that decreases the stability of p53 [25–27] , we examined the total protein levels of p53 in the UBE4B and LSD1 double-knockdown cells . We detected significantly increased p53 protein levels in the UBE4B and LSD1 double-knockdown cells when compared with the mock-knockdown control , indicating a stabilization of the p53 protein ( Fig . 4D ) . To further confirm that the single or double knockdowns of UBE4B and LSD1 were activating p53-mediated transcription , we used a firefly luciferase ( p53RE-luc ) reporter construct carrying p53-responsive elements in its promoter . Knockdown of either UBE4B or LSD1 increased p53 transcriptional activity . However , the simultaneous knockdown of both UBE4B and LSD1 resulted in an even higher p53 transcriptional activity ( Fig . 4E ) , consistent with the synergistic antiproteotoxicity effects of knocking down both UBE4B and LSD1 . To examine if the increased luciferase activity reflected p53-dependent transcriptional activation , we expressed MDM2 , an E3 ubiquitin ligase and negative regulator of p53 , or β-galactosidase as a control , together with the p53 activity reporter . The introduction of MDM2 significantly reduced p53-dependent transcriptional activation of the luciferase reporter under the UBE4B and LSD1 double-knockdown condition ( S4A Fig . ) , confirming the specificity of the regulation of p53 by UBE4B and LSD1 . LSD1 demethylates the p53 protein , and loss of LSD1 increases K370-p53 dimethylation , an activating post-translational modification specifically recognized by 53BP1 [28] . To determine whether the activation of p53 resulted partially from its enhanced interaction with 53BP1 , we co-immunoprecipitated p53 and 53BP1 from HEK293T cells in which LSD1 and UBE4B were previously knocked down . An increased amount of 53BP1 was pulled down by an equal amount of p53 protein in the double-knockdown cells when compared with the control , indicating an increased interaction between p53 and its coactivator , 53BP1 ( Fig . 4F ) . To confirm directly that the dimethylation on the p53 protein is increased , we performed western blotting using a dimethyl K370-p53-specific antibody , and detected an increase in the level of dimethylated p53 in the double-knockdown samples compared with the control and compared with the total p53 ( S4B Fig . ) . In sum , these data demonstrate that p53 , as a transcription factor , is significantly elevated and activated by the knockdown of UBE4B and LSD1 . Among the genes that were up-regulated by the UBE4B and LSD1 double knockdown in our microarray data set , there were a few that had been reported to be important for protein quality control , including forkhead box O3 ( FOXO3a ) , FOXO4 , and proteasome 26S subunit non-ATPase11 ( PSMD11 ) ( S3C Fig . and S2 Table ) . FOXOs are a family of transcription factors invoked in protein quality control [20 , 29 , 30] , and the 19S proteasome component PSMD11 is a critical regulator of proteasome activity [31 , 32] . It is notable that FOXO3a is transcriptionally up-regulated by p53 [33] , and PSMD11 is transcriptionally induced by FOXOs [31 , 32] . We confirmed through RT-qPCR that FOXO3a , FOXO4 , and PSMD11 were all transcriptionally up-regulated when UBE4B and LSD1 were knocked down ( Fig . 4G ) , linking these positive regulators of protein quality control downstream of p53 to the UBE4B- and LSD1-dependent antiproteotoxicity activity . Since FOXO3a is a transcription factor downstream of p53 and positively regulates protein quality control , we examined the FOXO3a-mediated transcriptional activity by employing a previously established luciferase reporter that is driven by the forkhead-responsive element ( FHRE ) [34] . A constitutively active form of FOXO3a , FOXO3a-TM , was co-expressed with the FHRE-luciferase reporter in HEK293T cells to measure the transcriptional activity of this particular FOXO member . Consistent with the p53 activation , we found that the FOXO3a activity is induced most strongly when both UBE4B and LSD1 are simultaneously knocked down ( Fig . 4H ) . Similar results were observed with another FOXO-family member , FOXO1 ( S4C Fig . ) . The daf-16 gene is the sole C . elegans ortholog of the mammalian FOXO family [35 , 36] . To examine a role of daf-16/FOXO in the SUNS pathway in C . elegans , we built a quadruple mutant strain carrying the SOD1G85R transgene , the SUNS mutations spr-5 ( by134 ) and ufd-2 ( tm1380 ) , and loss-of-function mutation daf-16 ( mu86 ) . In C . elegans locomotor assays , daf-16 ( mu86 ) partially blocked the rescuing effect of the SUNS mutations on SOD1G85R toxicity , suggesting that the daf-16 gene activity is partially required for the suppressor activity ( S4D Fig . ) . To determine whether the enhanced clearance of SOD1G85R upon the knockdown of UBE4B and LSD1 reflects an increase in proteasome-mediated degradation , we examined abundance of several proteasomal subunits . Consistent with our transcriptome analysis that demonstrated an increase in the level of PSMD11 RNA , the PSMD11 protein level was significantly increased in the double-knockdown cells ( Fig . 5A ) . Both PSMD11 and PSMD4 ( proteasome 26S subunit , non-ATPase , 4 ) are resident to the 19S regulatory proteasome particle , and we found that the PSMD4 protein level was also significantly increased . Moreover , components of the 20S core proteasome particle , including 20Sα3 , 20Sβ1 , and 20Sβ5 , were also significantly increased . To determine whether the increase in the quantities of proteasome subunits corresponds to augmented proteasomal activity , we measured the chymotrypsin-like proteasome activity in lysates derived from HEK293T cells with single or double knockdown of UBE4B and LSD1 using a luciferase assay ( S5A Fig . ) . The proteasomal activity was increased in LSD1 or UBE4B single-knockdown cells , but the highest activity was observed in double-knockdown cells ( Fig . 5B ) . This finding is consistent with the results of C . elegans suppressor studies and protein solubility assays in mammalian cells , in which both UBE4B and LSD1 are required for the maximal effects of the suppressors . Consistent with the activation of proteasomal activity by the double knockdown of UBE4B and LSD1 , inhibition of the proteasome by treating cells with MG132 blocked the degradation of SOD1G85R conferred by the UBE4B and LSD1 knockdown ( S5B Fig . and S5C Fig . ) . Thus , the knockdown of UBE4B and LSD1 significantly increases both the subunit quantity and the activity of proteasomes . In addition to the activation of the proteasome , we asked whether autophagy is also up-regulated in the UBE4B and LSD1 double-knockdown cells . To measure autophagic activity , we employed a Gaussia luciferase ( GLuc ) release assay [37 , 38] that reports the autophagy-dependent ATG4B cleavage of an actin-tethered actin-LC3-GLuc-fusion protein and its subsequent release from the cell into the medium ( S5D Fig . and S5E Fig . ) . HCT116 cells , which are amenable to this assay , were transfected with shRNA constructs to knock down UBE4B and LSD1 , with the LC3-GLuc plasmid used to measure the cleavage of LC3 and the constitutively secreted control ( cytomegalovirus-secreted embryonic alkaline phosphatase [CMV-SEAP] ) for transfection/secretion normalization ( S5D Fig . and S5E Fig . ) . The LC3-dependent GLuc activity , measured over a period of 72 h , showed a 2-fold increase in ATG4B proteolytic activity at the end of the time course demonstrating the activation of autophagy by the knockdown of UBE4B and LSD1 ( Fig . 5C ) . The cells transfected with the noncleavable , LC3-less fusion , the Act-GLuc construct , showed only background levels of Gluc activity , similar to the levels observed in nontransfected cells . To confirm that autophagy was activated by the UBE4B and LSD1 double knockdown , we measured LC3-II accumulation by western blotting in cells . Double knockdown samples showed increased levels of LC3-II compared with controls , even when LC3-II was elevated by treatment with 3-MA ( Fig . 5D ) . Together , these data demonstrate that a UBE4B- and LSD1-dependent protein quality control pathway similar to that in C . elegans also operates in mammalian cells , since a reduction in these two enzymes promotes the removal of aggregating proteins through enhanced post-translational quality control systems involving the proteasome and autophagy . Until now , p53 has not been associated with antiproteotoxicity activity . p53 has been shown to regulate autophagy , but in opposing directions [39] . Our microarray analysis and subsequent studies establish a correlation between the activation of p53-mediated transcription and enhanced protein quality control conferred by the knockdown of UBE4B and LSD1 ( Fig . 5 ) . It has been demonstrated that p53 is a target of polyubiquitination by UBE4B and demethylation by LSD1 , and each of these functions decreases the p53 activity [25 , 28] . Thus , p53 has emerged as a potential effector that mediates the synergistic action of UBE4B and LSD1 in the antiproteotoxicity pathway . To determine whether p53 directly protects against proteotoxicity , we first used small molecule activators of p53 in the cell-based SOD1G85R protein aggregation assay . Tenovin-1 activates p53 by inhibiting the SIRT1/2 deacetylase and therefore promoting p53 acetylation , thereby increasing its stability and activity [40] . CP-31398 is another drug that activates p53 by stabilizing the p53 DNA-binding domain in an active conformation and inhibiting its ubiquitination [41 , 42] . Both p53 activators significantly reduced the amount of SOD1G85R protein in the supernatant and pellet fractions ( Fig . 6A , S6A Fig . , and S6D Fig . ) at various concentrations of Tenovin-1 ( 0 . 4–1 . 2 μM ) and CP-31398 ( 2–4 μg/ml ) . Furthermore , to confirm that Tenovin-1 and CP-31398 acted through the activation of p53 , we reduced p53 levels via shRNAs while testing the clearance of SOD1G85R protein with the drug treatment . The p53 reduction blocked the effect of either Tenovin-1 or CP-31398 on promoting SOD1G85R clearance , indicating that the drugs act through p53 ( S6G Fig . ) . To investigate the mechanism by which the activation of p53 by Tenovin-1 and CP-31398 reduced the levels of SOD1G85R proteins , we asked whether autophagy was activated by these drug treatments . In agreement with a previous report that CP-31398 activates autophagy [43] , we observed that increasing concentrations of either Tenovin-1 or CP-31398 up-regulated LC-II protein levels ( S6H Fig . ) , consistent with the activation of autophagy . To further validate that Tenovin-1 and CP-31398 act post-translationally to promote the clearance of SOD1G85R proteins , we performed cycloheximide chase experiments and confirmed the increase in the degradation of SOD1G85R proteins upon treatment of either of the p53 activator drugs ( S6B Fig . , S6C Fig . , S6E Fig . , and S6F Fig . ) . Conversely , we asked whether reducing p53 activity would negatively affect the clearance of SOD1G85R proteins . First , we performed the SOD1G85R protein solubility assay in cells in which p53 was reduced by RNAi . We found that partial removal of p53 in HEK293T increased SOD1G85R protein levels in both the supernatant and pellet fractions ( Fig . 6B ) . Next , using a human HCT116 cell line in which p53 was knocked out , we asked how the complete removal of p53 affected the clearance of misfolded SOD1G85R . Unlike the WT SOD1 protein , whose level was not affected by the absence of p53 , the SOD1G85R mutant protein was significantly increased in the p53 knockout cells when compared with the controls , indicating that endogenous p53 promotes the clearance of misfolded proteins ( Fig . 6C ) . To determine whether p53 mediates the UBE4B- and LSD1-dependent clearance of the SOD1G85R proteins , we knocked down UBE4B and LSD1 with or without the removal of p53 and then examined the levels of SOD1G85R . We applied both transient and stable shRNA knockdown by creating an inducible , stable HEK293T cell line expressing tetracycline-regulated shRNAs against UBE4B , LSD1 , and p53 . Both transient and stable knockdown of p53 significantly reversed the SOD1G85R protein clearance conferred by the UBE4B and LSD1 knockdown ( Fig . 7A and S7A Fig . ) . This result was confirmed with an independent set of shRNAs against UBE4B and LSD1 ( S7B Fig . ) and by knockdown of UBE4B and LSD1 in HCT116 cells with or without the p53 gene knocked out ( S7C Fig . ) . Notably , unlike HEK293T cells , HCT116 cells do not express SV40 large T-antigen ( LT-Ag ) , a regulator of p53 stability [44] , suggesting that the action of UBE4B and LSD1 does not require LT-Ag . Taken together , these results demonstrate that p53 is required for the UBE4B- and LSD1-dependent clearance of the SOD1G85R proteins , and it acts downstream of UBE4B and LSD1 to positively regulate the clearance of misfolded proteins . To confirm that p53 can modulate proteotoxicity in vivo , we used the Drosophila TDP-43M337V neurodegeneration model as described earlier ( Fig . 2D ) [22] . Either knockdown of p53 by RNAi or expression of a dominant negative form of Drosophila p53 ( p53 . R155H ) [45] with the GMR-Gal4 driver exacerbated the TDP-43M337V-induced eye phenotype ( Fig . 7B ) . This aggravation of the phenotype was evident in increased loss of pigmented ommatidia and , in p53 RNAi flies , the appearance of necrotic patches , which were observed at low penetrance ( Fig . 7B ) . Expression of the dominant negative p53 . R155H transgene on its own , in a wild-type background , did not cause any eye phenotype . Together , these results indicate that endogenous p53 plays a role in reducing TDP-43M337V proteotoxicity in the Drosophila eye . We further tested whether p53 activation would alleviate SOD1G85R-induced neurotoxicity . We employed a previously characterized SOD1 neurotoxicity assay [46] , in which spinal cord primary motor neurons were prepared from rat embryos , maintained on astrocyte monolayers supplemented with neurotrophic factors , and stained with a mature motor neuron marker , the neurofilament H ( NF-H ) antibody SMI-32 . Expression of SOD1G85R via a neuron-specific herpes simplex virus ( HSV ) vector reduced the survival of motor neurons significantly by approximately 50% in contrast to the HSV-LacZ control over 5 days ( Fig . 7C ) . When treated with the p53 activator Tenovin-1 , the motor neurons showed protection from SOD1G85R-induced proteotoxicity , as compared with the vehicle control . Among the various concentrations tested , 0 . 8 μM of Tenovin-1 completely blocked the neurotoxicity of SOD1G85R , with minimal toxicity from the drug itself ( Fig . 7C ) . These results confirm that the activation of p53 provides protection against the toxicity of misfolded proteins in neurons .
We isolated the SUNS C . elegans mutant based on the potent suppression of SOD1-induced neurotoxicity . The suppressor was found to significantly enhance the removal of misfolded proteins , underscoring the critical role of protein misfolding in SOD1-mediated neurodegeneration . The enhanced clearance also applies to other misfolded proteins , such as TDP-43 , FUS , and polyglutamine-containing proteins , indicating a general improvement in protein quality control . This rare but strong suppressor requires modulation of only two genes , suggesting that it provides a major protein quality control program with a readily accessible switch . Furthermore , the synergistic cooperation of two genes , ufd-2 and spr-5 , points to a common downstream pathway with integrative regulation . Consistent with the observation that the loss of function of C . elegans ufd-2 and spr-5 promotes the clearance of misfolded proteins , inactivation of their Drosophila and mammalian orthologs reduces the toxicity of aggregation-prone proteins , indicating the existence of a protein quality control regulatory mechanism that is functionally conserved across species . Interestingly , both mammalian genes encode lysine-modifying enzymes: UBE4B is a U-box type ubiquitin ligase , and LSD1 is a lysine-specific protein demethylase . Both UBE4B and LSD1 are highly expressed in neurons and essential for early development in mammals [47–50] . In contrast to the conventional notion that ubiquitin ligase promotes protein degradation , our studies indicate that UBE4B negatively affects the clearance of misfolded proteins , and its down-regulation protects against severe proteotoxicity . In line with our observation that the down-regulation of UBE4B protects against proteotoxicity in the nervous systems of C . elegans and Drosophila , mice with elevated levels of UBE4B show autophagy defects with accumulation of ubiquitin- and p62-positive aggregates in the brain [51] . UBE4B forms a complex with an AAA-ATPase p97/VCP to ubiquitinate and degrade specific client proteins [47 , 52] . p97/VCP plays an essential role in handling unfolded proteins , such as in endoplasmic-reticulum-associated protein degradation [53] , and it was recently linked to familial ALS [54] . Our findings thus provide a new link between p97/VCP and protein quality control , which is regulated by UBE4B . The fact that both UBE4B and LSD1 are enzymes catalyzing post-translational modifications suggests that their effects on protein quality control can be timely , energy-efficient , and integrative . The synergistic interaction between the two lysine-modifying enzymes , UBE4B and LSD1 , also suggests that their downstream pathways converge to influence protein quality control . Consistent with recent studies showing enhancement of protein quality control [20 , 55–57] , the identification of the strong antiproteotoxic effects mediated by UBE4B and LSD1 demonstrates that plasticity of the cellular protein quality control programs can be substantially augmented to yield overall protection to an organism . Unbiased transcriptome analysis points to p53 as a central regulator of the transcriptional reprograming that mediates the effects of UBE4B and LSD1 on protein quality control . Consistent with this observation , p53 has been found to have a number of direct transcriptional targets functioning in protein quality control and neuroprotection , and it also activates additional stress-response transcription factors such as FOXOs [33] . Interestingly , p53 is elevated in the central nervous system of patients with neurodegenerative conditions such as Alzheimer disease and ALS [58 , 59] . Our observation that the transcription factors mediate the effects of this strong suppressor is reminiscent of other signaling pathways governing protein homeostasis . For example , the heat shock response activates the expression of molecular chaperones and other protein quality control machinery via the master transcription regulators , the heat shock factors [60] . Also , the unfolded protein response promotes the endoplasmic reticulum ( ER ) quality control programs through the activation of a set of the transcription factors , including XBP1 , ATF4 , and ATF6 [61] . In recurring themes , the post-translational regulation by UBE4B and LSD1 activates the p53 transcription factor , which is then capable of eliciting a systematic protective program against proteotoxic stress . p53 has a well-established role in regulating responses to DNA damage [62 , 63] , and recently , a neuroprotective role of an activated DNA damage checkpoint has been demonstrated in a tau-dependent neurodegeneration model [64] . Here we propose that p53 is a versatile transcriptional switch that guards against both genotoxicity and proteotoxicity . The specific activity of p53 may be fine-tuned at the post-translational level by upstream regulators such as UBE4B and LSD1 . In addition , it is known that p53 promotes apoptosis in cells with irreversible genotoxic damage [65] . p53 may also function as a dual regulator in proteotoxicity: it promotes the repair and survival of moderately damaged cells but turns on cell death pathways in cells whose damage is irreparable . Such duality has been observed for other protein quality control systems , such as the ER stress responses [61] . Thus , p53 could serve as a critical regulator of cellular responses to proteotoxicity by repairing or removing damaged cells . Taken together , these findings reveal a previously unrecognized pathway that systematically antagonizes the proteotoxicity associated with neurodegenerative diseases , and they point to potential targets for harnessing the protective capacity of the cells’ reprogrammed protein quality control to develop a wide-spectrum antiproteotoxicity therapeutic strategy .
The pregnant rat dams were euthanized by overdosing with nembutal . The Children’s Hospital of Philadelphia IACUC approved these procedures ( protocol #597 ) . For mammalian expression , SOD1 and TDP-43 were expressed in pEF-BOS and pRK5-Myc , respectively , as previously described [9 , 20] . The UBE4B ( TF308519 ) and LSD1 shRNA ( TF316984 ) plasmids and the scrambled control ( TR30015 ) were from Origene . The p53 shRNA plasmid pLVTH-sip53 and control pLVTH were from D . Trono ( Addgene #12239 ) [66] . The p53 transcriptional reporter PG13-Luc was a generous gift from B . Vogelstein [67] . The autophagy luciferase release plasmids Act-LC3-Gluc and Act-Gluc were kindly provided by B . Seed [37] , and the control pCMV-SEAP was from A . Cochrane ( Addgene #24595 ) . For transgenic C . elegans , ufd-2 and spr-5 complementary DNAs ( cDNAs ) were cloned into a vector under the control of an snb-1 promoter , as previously described [20] . Additional information on the shRNA targeting sequences and vectors is given in the Supporting Information Materials and Methods ( S1 Text ) . The Bristol N2 C . elegans strain was used in all experiments unless otherwise specified . A list of C . elegans strains is given in the Supporting Information Materials and Methods ( S1 Text ) . Transgenic lines were generated according to standard procedures by injecting 20 μg/ml of expression plasmid DNA into hermaphrodite gonads . For the suppressor screen , worms were mutagenized with 47 mM ethyl methanesulfonate , and a semiclonal strategy was used with five P0 worms in one plate . Suppressors were visually selected based on strong recovery in the movement phenotype in the F2 generation . The suppressor mutations were mapped by using single-nucleotide polymorphism markers in the Hawaiian strain and then identified by whole-genome deep sequencing , followed by Sanger sequencing validations ( see Supporting Information Materials and Methods [S1 Text] ) . The C . elegans strains were observed stereoscopically , and their motility was quantified by the thrashing assay [20] . Animals were transferred from the feeding plate into M9 buffer ( 3 mg/ml KH2PO4 , 6 mg/ml Na2HPO4 , 5 mg/ml NaCl and 1 mM MgSO4 ) . After 1 min of adaptation , the number of body bends or thrashes was counted for 1 min as an index of the locomotor phenotype . A thrash was counted when both the head and the tail bent away from the anteroposterior axis by more than 45° . Videos of C . elegans locomotion were recorded using a Leica M165 fluorescence stereoscope . High-magnification imaging was performed on a Zeiss AxioObserver Z1 with Apotome , with C . elegans immobilized by 10 mM levamisole . See Supporting Information Materials and Methods ( S1 Text ) . See Supporting Information Materials and Methods ( S1 Text ) . The protein solubility assay to measure aggregate proteins in C . elegans and mammalian cells was modified from a previously described protocol [9] ( see Supporting Information Materials and Methods [S1 Text] ) . After a 72-h knockdown , cells were detached and transfected with firefly luciferase p53 reporter plasmid ( PG13-luc ) , together with a thymidine kinase promoter Renilla luciferase ( tk-Rluc ) reporter for normalization . Cells were lysed in passive lysis buffer ( Promega ) 24 h after transfection and analyzed with the Dual Luciferase Reporter System according to the manufacturer’s recommendations ( Promega ) using an injector-equipped Synergy H1 microplate reader ( Bio-Tek ) . Proteasome assays were performed as described previously [68] , using the Suc-LLVY-Luciferin substrate for chymotrypsin-like activity of the proteasome ( the Proteasome Glo kit , Promega ) . In brief , cells were detached and washed in DMEM/10 , followed by several washes in cold PBS . Proteasome lysis buffer ( 50 mM Tris-HCl , pH 7 . 5 , 0 . 025% digitonin , 250 mM sucrose , 5 mM MgCl2 , 0 . 5 mM EDTA , 2 mM ATP , and 1 mM DTT ) was added to the cells and incubated on ice for 5–10 min . The lysates were then centrifuged for 15 min at 20 , 000 g to isolate the cytoplasm containing the proteasomes . The supernatant was transferred to a fresh tube , and equal amounts of protein were used in each assay . Autophagy was quantified with a Gaussia luciferase release assay [37 , 38] , which is based on the ATG4B-induced proteolytic cleavage of an actin-anchored fusion LC3-Gluc fusion protein ( S5D Fig . and S5E Fig . ) . ATG4B-induced proteolytic cleavage of LC3 releases the Gluc fragment and enables its secretion into the cell medium . The activity of the released Gluc in the medium ( together with constitutively secreted SEAP ) was measured by the Secrete-Pair Dual Luminescence Assay kit ( GeneCopoeia ) . Cells , plated in 12-well dishes , were transfected with the Act-LC3-Gluc or control Act-Gluc plasmid together with the normalization control , CMV-SEAP . At 24 h after transfection , the DMEM/10 medium was replaced , and 100 μl of cell growth medium was withdrawn at 24 h , 48 h , and 72 h . The medium was centrifuged at 6 , 000 g for 5 min to remove detached cells , followed by the luciferase analysis according to the manufacturer’s recommendations ( GeneCopoeia ) using a microplate reader ( Synergy H1 , Bio-Tek ) . For LC3 western blot analysis , cells were lysed in LC3 buffer ( 50 mM Tris-Cl , pH 8 . 0 , with 1% SDS , 0 . 5% NP40 , 150 mM NaCl , and 5 mM EDTA ) and sonicated with a Diagenode Bioruptor device ( set on high , 30-sec pulse , 30-sec pause , 7 . 5 min total ) . Total RNA was isolated from HEK293T cells with the RNeasy Mini kit and analyzed using the Affymetrix human GENE 1 . 0ST array . The microarray data were managed using the Partek Genomic Suite ( Partek , St . Louis ) and Spotfire DecisionSite software ( TIBCO Software , Palo Alto , California ) and analyzed using Ingenuity Pathways Analysis software ( IPA , Ingenuity Systems ) . In addition to RNA , total protein was isolated from the same samples by acetone precipitation and resolubilizing the flow-through lysates to verify the reduction of the UBE4B and LSD1 proteins . For quantitative RT-qPCR validations , cDNAs were synthesized with the QuantiTect reverse transcription kit ( Qiagen ) . Primers for quantitative RT-qPCR were from PrimerBank ( S3 Table ) [69] . RT-qPCRs were performed on a BioRad thermal cycler with iQ SYBER Green PCR mix ( BioRad ) . Embryonic Sprague Dawley rat spinal cord cultures and neuronal survival assays were previously described [46] ( see Supporting Information Materials and Methods [S1 Text] ) . The p-values for all analyses were obtained using Student’s t tests performed in Excel or GraphPad Prism 6 , unless otherwise indicated . For the microarray data , Student’s t test was used to analyze the gene expressions . For the Upstream Regulator Ingenuity Pathway Analysis , Fisher’s exact test was used . For the LC3-II western blotting analysis and the spinal cord motor neuron toxicity assay , a one-way ANOVA with multiple comparison test was used . | To function properly , proteins must assume their correct three-dimensional shapes . There are numerous mechanisms within the cell , collectively referred to as protein quality control ( PQC ) , that verify proper folding . If abnormal folding is detected , PQC can either help the protein to refold or target it for degradation . Failures in protein folding and PQC lead to the accumulation of misfolded proteins , which often self-associate into large aggregations that are thought to be the underlying cause of several neurodegenerative diseases . In this study , we use the roundworm Caenorhabditis elegans as a model to understand how cells handle disease-associated misfolded proteins . In a large-scale genetic screen , we discovered two suppressor genes , ufd-2 and spr-5 , which encode a ubiquitin ligase and a lysine-specific demethylase , respectively . When these two proteins are inactivated , we observe a marked reduction in the toxicity of several misfolded proteins . ufd-2 and spr-5 are conserved in humans ( UBE4B and LSD1 , respectively ) , as are their effects on misfolded proteins . We show that UBE4B and LSD1 regulate the activity of protein degradation machineries including the proteasome and autophagosomes . Using microarrays and biochemical analyses , we identify p53 as a key downstream transcription factor that mediates the action of UBE4B and LSD1 on protein clearance . This work establishes p53 as a regulator of proteome integrity and uncovers a new protein quality control pathway that could potentially be exploited to increase the degradation of misfolded proteins in diseased cells . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2015 | Regulation of Protein Quality Control by UBE4B and LSD1 through p53-Mediated Transcription |
Bacteria are able to sense and respond to a variety of external stimuli , with responses that vary from stimuli to stimuli and from species to species . The best-understood is chemotaxis in the model organism Escherichia coli , where the dynamics and the structure of the underlying pathway are well characterised . It is not clear , however , how well this detailed knowledge applies to mechanisms mediating responses to other stimuli or to pathways in other species . Furthermore , there is increasing experimental evidence that bacteria integrate responses from different stimuli to generate a coherent taxis response . We currently lack a full understanding of the different pathway structures and dynamics and how this integration is achieved . In order to explore different pathway structures and dynamics that can underlie taxis responses in bacteria , we perform a computational simulation of the evolution of taxis . This approach starts with a population of virtual bacteria that move in a virtual environment based on the dynamics of the simple biochemical pathways they harbour . As mutations lead to changes in pathway structure and dynamics , bacteria better able to localise with favourable conditions gain a selective advantage . We find that a certain dynamics evolves consistently under different model assumptions and environments . These dynamics , which we call non-adaptive dynamics , directly couple tumbling probability of the cell to increasing stimuli . Dynamics that are adaptive under a wide range of conditions , as seen in the chemotaxis pathway of E . coli , do not evolve in these evolutionary simulations . However , we find that stimulus scarcity and fluctuations during evolution results in complex pathway dynamics that result both in adaptive and non-adaptive dynamics depending on basal stimuli levels . Further analyses of evolved pathway structures show that effective taxis dynamics can be mediated with as few as two components . The non-adaptive dynamics mediating taxis responses provide an explanation for experimental observations made in mutant strains of E . coli and in wild-type Rhodobacter sphaeroides that could not be explained with standard models . We speculate that such dynamics exist in other bacteria as well and play a role linking the metabolic state of the cell and the taxis response . The simplicity of mechanisms mediating such dynamics makes them a candidate precursor of more complex taxis responses involving adaptation . This study suggests a strong link between stimulus conditions during evolution and evolved pathway dynamics . When evolution was simulated under conditions of scarce and fluctuating stimulus conditions , the evolved pathway contained features of both adaptive and non-adaptive dynamics , suggesting that these two types of dynamics can have different advantages under distinct environmental circumstances .
Bacterial responses to external stimuli and the pathways with which they are mediated are model systems for studying the molecular basis of behaviour . Much of the research in this field has focused on chemotaxis , the ability of bacteria to swim up a gradient of a chemical attractant , as performed in the model organism Escherichia coli . More than 30 years after the first studies [1] , [2] , we now have extensive knowledge of the underlying biochemical pathway [3] , [4] . Briefly , E . coli swims in a forward direction ( undergoing some degree of rotational diffusion ) when the reversible motor proteins on its outer membrane rotate counter-clockwise ( CCW ) and the attached flagella intertwine to form an effective propeller . When the motors reverse and rotate clockwise ( CW ) , the flagella disassociate and cause the bacterium to tumble , resulting in a new swimming direction . The switching frequency of the motor is coupled to receptor activity by a set of proteins constituting a signalling pathway . With increasing attractant levels , the excitatory branch of the pathway causes suppression of CW rotation and tumbling , while the adaptation branch causes the cell to resume its original tumbling levels at constant attractant concentrations independently of this concentration level . The former branch involves the receptor-coupled kinase CheA and the associated response regulator CheY , which when phosphorylated binds the motor and increases the probability of CW rotation . The adaptation is achieved via control of receptor methylation , and hence receptor sensitivity , through the proteins CheR and CheB . The combination of these two branches results in the tumbling frequency approximately following a negative time-derivative of the attractant concentration [5] . Adaptation is the hallmark of this response , allowing bacteria to perform temporal comparisons of attractant with high sensitivity over a wide dynamic range [5]–[9] . While this ‘E . coli paradigm’ of chemotaxis is well established , our knowledge of taxis responses towards other stimuli and in other species [10]–[12] indicates that the derivative response with adaptation observed in E . coli is neither universal nor necessary for effective taxis . For example , the response to oxygen in E . coli is believed to be mediated by the receptor Aer that lacks methylation sites , indicating lack of adaptation [13] . In Rhodobacter sphaeroides , adaptation to persistent stimuli occurs much slower or not at all [14] . In the same species , growth under aerobic conditions results in an ‘inverted’ chemotaxis response where increasing attractant concentration causes an increase in tumbling frequency [15]; this inverted response is also observed in certain Halobacteria [16] and in certain mutant strains of E . coli that have been ‘gutted’ of some or most of the chemotaxis proteins [17] , [18] . Interestingly , these natural and mutant strains all still show the ability to chemotax . These diverse chemotaxic dynamics could result from multiple pathways that allow bacteria to integrate information about the internal and external state to produce coherent taxis behaviour [13] . For instance , in R . sphaeroides , genetic studies indicate that the large number of taxis proteins in this species are arranged into several distinct pathways [19] . The chemotaxic response towards certain molecules in this species requires transport into the cell [20] , demonstrating the link between metabolism and chemotaxis . Despite much effort , we still lack a comprehensive understanding of the different molecular mechanisms involved in bacterial taxis responses , their underlying dynamics , and how these different dynamics are integrated . Here , we use a computational approach to address some of these questions by simulating the evolution of a taxis response using computer modelling of bacterial movement and pathway dynamics . These simulations use a population of virtual bacteria existing in a virtual world complete with a stimulus source that is assumed to signal the presence of favourable conditions . Bacteria start with a set of non-interacting proteins , as well as a receptor and a reversible motor . Interactions between the proteins evolve through random mutations , with bacteria selected for reproduction based on their ability to localise at sites of favourable conditions . These evolutionary simulations consistently result in bacteria with a strong ability to move towards the stimulus . Interestingly , under conditions of abundant stimuli these bacteria evolve so that the tumbling probability is directly coupled to stimulus levels without any adaptation . We find that such non-adaptive dynamics can be mediated by as few as two signalling components , allowing for the possibility of metabolites acting as effectors . Simulation conditions mimicking environments with scarce or fluctuating stimulus sources result in evolution of pathways with complicated dynamics that have features of both non-adaptive and adaptive dynamics . Combined with experimental observations , these results demonstrate that adaptive dynamics are not necessary for effective chemotaxis . This work also suggests that non-adaptive dynamics underlie the chemotaxis observed in gutted E . coli strains and may have a role in the complex taxis behaviour of R . sphaeroides . We speculate that mechanisms leading to such dynamics exist in current-day bacteria and provide a way to fine-tune taxis responses in different conditions , or link it to the energy state of the cell .
In order to study the evolution of bacterial taxis , we use virtual bacteria that move in a computer-based two-dimensional environment containing a fixed stimulus source and periodic boundary conditions . As such , this computer environment mimics a natural environment with abundant stimuli . The movement of these bacteria is coupled to the dynamics of a signalling pathway consisting of several proteins that catalyse each other's activation and deactivation , corresponding to kinases and phosphatases in real cells ( see Methods ) . These proteins include a receptor whose activity level is directly coupled to the local stimulus level and an effector that , when activated , can bind to a reversible motor , reversing its direction and causing the bacteria to tumble . Evolutionary simulations start with a population of bacteria , each of which contains a certain number of proteins that are initially non-interacting . These bacteria are allowed to explore the environment for a ‘generation-time’ consisting of a certain number of time steps . At each time step , the bacteria either can continue to swim forward or can tumble to orient to a new random direction . Additionally , the concentrations of activated proteins in the pathways of each bacterium are updated , and the probability of tumbling during the next time step is computed based on the concentration of activated effector . After this generation-time , bacteria are selected for replication based on the integrated amount of stimulus they have encountered . During replication , there is a probability for mutations to occur , which alters the structure and parameters of the biochemical pathway . To summarise , these evolutionary simulations couple mutational events occurring at molecular level ( i . e . pathway level ) with selection at behavioural level ( i . e . taxic response ) . Note that this is a generic model where the stimulus represents anything capable of activating a receptor ( e . g . , chemical attractant , light , pH ) and there are no a priori assumptions regarding pathway structure or dynamics . Figure 1 shows the population average of fitness ( encountered stimuli ) during one evolutionary simulation for a signalling pathway consisting of four proteins . As shown , the fitness value rapidly improves over a few generations and reaches a plateau . Clearly , the pathway structure and dynamics in virtual bacteria are evolving in such a way to mediate taxis . This behaviour can be seen from the average time spent by the population at different parts of the environment ( see insets of Figure 1 ) . While un-evolved bacterial populations are distributed irrespective of stimulus source , final populations are able to quickly co-localise with it . This behaviour is mediated by a specific biochemical pathway dynamics; at steady state , in absence of any signal , the concentration of activated effector is at a low level and the bacterium mostly swims without tumbling ( see Figure 2 for typical pathway structure and dynamics , kinetic parameters are shown in Dataset S1 ) . When the bacterium encounters higher stimulus levels , the effector is rapidly activated and stays activated as long as the signal is present , resulting in increased bacterial tumbling . We find that the qualitative nature of this type of dynamics is independent of basal stimuli level ( data not shown ) . This non-adaptive dynamics allow the bacteria to spend more time in regions of high stimulus and swim straight when the stimulus level decreases . In evolutionary simulations repeated five times for pathways of 2 to 5 proteins , this mechanism always evolved as the dominant one . The structures of pathways resulting from these simulations were diverse ( see Dataset S1 ) indicating that there are several possible biochemical signalling cascades that can mediate non-adaptive dynamics . In case of the sample pathway shown in Figure 2 , we find that the receptor acts as a global inhibitor shutting down effector activity in absence of stimuli . Incoming signals suppress receptor activity , allowing a build up of effector , which is involved in a feedback loop with one of the intermediary proteins , protein one ( see cartoon representation in Figure 2 ) . The other protein acts as a kinase ( i . e . activator ) on both the receptor and protein one , thereby ensuring rapid response termination when the stimulus is removed . This complex pathway structure and the resulting dynamics allow efficient chemotaxis behaviour as described above . However , similar dynamics can be achieved with much simpler circuits containing only two proteins ( see Discussion ) . Using a simple analytical model , we can capture the movement of bacteria as mediated by non-adaptive dynamics ( see Methods ) . This model shows that in simple environments the presented dynamics should lead to bacteria accumulating approximately proportionally with the local level of the stimulus . Note that as long as the stimulus levels are above a certain threshold , this mechanism is only sensitive to the ratio of the relative levels and not their absolutely magnitudes . This suggests that an efficient taxis response can be achieved over a wide dynamic range of stimuli with pathway-dynamics that does not display adaptation to stimulus and results in increases in tumbling probability with increasing stimulus . Both these dynamical features are in striking contrast to the chemotaxis behaviour of E . coli , where the pathway ensures decreasing tumbling probability with increasing stimulus followed by rapid adaptation [5] ( see Figure 2 ) . There are a number of different possible explanations for why the taxis pathways evolved in these simulations are characterised by an ‘inverted’ response ( i . e . response to increasing stimuli is opposite of that seen in E . coli ) and non-adaptive dynamics . Firstly , the evolutionary processes as modelled here might make non-adaptive pathways more evolutionarily accessible . Secondly , it might be that the modelled environmental situations are particularly well-suited for taxis mediated via such dynamics . In particular , the assumptions of stimulus consistency and abundance in the environment might reduce the need for adaptation . Thirdly , there could be other factors such as intra-cellular communication , multi-state receptors ( i . e . receptors with methylation sites ) , and various physical processes [21] that are not included in the model and that could be important for the evolution of taxis responses mediated by other dynamics . To see if the non-adaptive dynamics were the result of the difficulty of evolving adaptive dynamics , we performed additional simulations . These started with an initial bacterial population containing biochemical pathways with dynamics similar to that found in E . coli [22] . In five separate simulations , the bacteria always evolved more fit pathways with non-adaptive dynamics and inverted response . In other words , under the conditions of these simulations ( i . e . under high stimulus abundance ) , there always existed a pathway with non-adaptive dynamics that could mediate a more efficient taxis response than the original adaptive pathway . This indicates that the results we obtain are not due to lack of an evolutionary route to the conventional dynamics observed in E . coli . It does not indicate , however , that taxis responses mediated by non-adaptive dynamics are superior as it was not possible to reproduce all environmental conditions and the other possibly-important features as mentioned above . To explore the effect of environmental conditions on the evolution of chemotaxis , we ran two sets of simulations under ( i ) periodic-boundary conditions and fluctuating stimulus source and ( ii ) non-periodic boundary conditions and fixed stimulus source . The first set of conditions allow us to test the hypothesis that adaptive dynamics provide a means for bacteria to preserve robustness of the response to fluctuations in the external environment or internal parameters [23] . The latter conditions mimic an environment with scarce stimulus , where exploration is expected to be more important than exploitation . In five simulations run under each condition , we did not find pathways with dynamics that are adaptive over a wide range of stimuli as seen in E . coli . However , several simulations resulted in pathways that had dynamical behaviour similar to that of E . coli under some conditions . As shown in Figure 3 , these pathways give a “normal” response ( i . e . decreased tumbling probability with stimuli ) and have limited adaptation to continuous stimuli . Interestingly , most pathways evolved under non-periodic boundary conditions show dynamics that are dependent on basal stimuli levels . This affects mostly the adaptation dynamics and we observe one pathway achieving perfect adaptation under a narrow range of basal stimuli levels ( see Figure 3 ) . Most simulations run under sparse stimulus conditions resulted in approximately same fitness levels as shown in Figure 1 . However , simulations run under these conditions ( ii ) took much longer ( usually more than 2000 generations ) to reach these fitness levels . Taken together , these results indicate that realistic and complex environmental conditions lead to evolution of complex pathway dynamics that contain features of both adaptive and non-adaptive dynamics . Untangling the role of each type of dynamics in the efficiency of chemotaxis requires further detailed analyses .
The molecular systems mediating the taxis responses observed in bacteria are more complicated than the dominant picture of E . coli chemotaxis suggests . Bacteria can sense and respond to a variety of environmental clues , possibly integrating the signal from different biochemical pathways . Recent experimental observations from an increasing number of bacterial species and past studies from mutant strains of E . coli hint at the diversity of molecular mechanisms involved in generation of these responses . Here we provide evidence for the effectiveness of one possible dynamical scheme . The main features of this dynamics is an inverted response , leading to increasing tumbling frequency with increasing stimulus level , and an absence of adaptation to continuous stimuli . We show that such non-adaptive dynamics readily evolve under different environmental conditions and model assumptions and allow bacteria to accumulate at favourable conditions efficiently . These findings provide a possible explanation for the non-adaptive dynamics and inverted responses observed in wild type R . sphaeroides [14] , [15] and the inverted responses observed in Halobacteria [16] and gutted strains of E . coli [18] . In each case , efficient taxis responses were observed , although the exact nature of the underlying molecular mechanisms could not be determined . It is likely that these mechanisms form systems similar to the pathways presented here . An analysis of results from simulations with two proteins reveals the minimum signalling systems to achieve taxis responses mediated by non-adaptive dynamics ( see Figure 4 ) . They involve coupling of the signal to an effector via a receptor , with self-regulation of both proteins ( through allosteric interactions or processes such as auto-phosphorylation ) . The striking simplicity of these minimal systems lead to the speculation that non-adaptive dynamics could even be achieved without any signalling proteins; a small molecule , that is a by-product of metabolism or is taken into the cell via a transporter , could directly regulate tumbling probability of the cell . We hypothesise that exactly such a scenario is responsible for chemotaxis observed in gutted E . coli [18] . Attractant-related metabolism causes increases in fumarate levels inside the cell , which binds the motor and increases tumbling probability . It has been demonstrated experimentally that fumarate can be involved in chemotaxis [24] and can control motor switching [25] , although the exact dynamics of how it could lead to chemotaxis was unknown . If non-adaptive dynamics are available and provide efficient taxis responses , why do we observe adaptive dynamics in E . coli and other bacterial species ? Adaptive mechanisms might be more efficient in exploring the environment and achieving a robust response under fluctuating stimuli . The evolution of complex dynamics in simulations run under conditions mimicking scarce and fluctuating stimulus sources supports such arguments . The dynamics of these pathways contained both adaptive and non-adaptive features , further indicating the possible complexity of chemotaxis behaviour . As indicated by experiments with gutted E . coli and other species , non-adaptive pathways probably function in conjunction with the canonical mechanism , and are involved in the fine-tuning of taxis responses under certain environmental conditions or in providing a link between energy related taxis responses and chemotaxis . Alternatively , given the simplicity of the required molecular machinery , the non-adaptive dynamics could be the precursor of the more complicated adaptive mechanisms . Both hypotheses could be tested with specific experimental setups and a sequence analysis of the proteins involved in taxis responses respectively .
To study the pathways underlying taxis responses we used a previously described pathway model [22] . This model assumes that a pathway consist of a set of Np proteins , all of which can exist in a deactivated or activated state ( activation can correspond to phosphorylation , methylation , or any other type of chemical or structural modification ) . Each protein is capable , in the activated state , of causing the activation or deactivation of any of the other proteins . The first protein in the pathway is arbitrarily chosen to act as a receptor that can be activated by the external stimuli , while protein Np is arbitrarily chosen to be an effector . When activated , it can bind to the motor protein causing a reversal of the motor and the bacterium to tumble . The biochemical dynamics for [Pi*] , the fraction of protein i that is activated , obeys ( 1 ) where kij ( lij ) is the rate at which activated protein j activates ( deactivates ) protein i , kii ( lii ) is protein i's rate of self-activation ( deactivation ) , k1A ( l1A ) represents the rate at which the stimulus activates ( deactivates ) the receptor protein 1 , [A] is the local level of stimulus , and δ is a Kronecker delta . The total concentration of each protein is assumed to be one . For simplicity we assume that any given protein can either activate or deactivate another , but not both . That is , kij lij = 0 for all i and j . We formulate this by considering a value of γij , where positive ( negative ) values of γij , correspond to positive values of kij ( lij ) . This can be expressed as ( 2 ) where H ( x ) is the Heaviside step function , equal to one if x is positive , and zero otherwise . The probability pTumble of a protein tumbling at any given timestep is given by ( 3 ) where γm is the affinity of protein Np for the motor . The evolutionary simulations are carried with a population of Nbacteria = 1000 bacteria existing on a two-dimensional 100×100 square space with periodic boundary conditions . The level of stimulus is defined as a Gaussian distribution located at the centre ( 50 , 50 ) of the space with maximum value 10 . 0 and width 7 . 1 . At each generation , every bacterium starts at a location ( 30 , 30 ) pointed in a random direction , with all activated protein concentrations set to zero . At each timestep the bio-kinetic equations ( Equation 1 ) are integrated based on the local level of stimulus . The bacterium then can either tumble ( choose a new random direction ) with probability pTumble or , alternatively swim in a straight line in the current direction a distance given by the swimming speed ( 0 . 2 ) . Each bacterium absorbs an amount of stimulus proportional to the local stimulus level , without consuming it . At the end of Nstep = 5000 timesteps , the next population of bacteria are selected using tournament selection: a set of five bacteria is chosen at random , and of the five , the bacterium that has accumulated the most stimuli is assigned to the next generation . This procedure is repeated ( with replacement ) Nbacteria times . During each assignment process , there is a probability pmutation = 0 . 1 of a mutation occurring . This mutation involves adding or subtracting a Gaussian-distributed value ( mean 0 , standard variation 0 . 1 ) to one of the γij chosen at random . In summary , the resulting pathway model captures the basic biochemistry of signalling pathways and allows coupling of external signals to the tumbling probability . The basic assumptions of the model are that proteins can only occur in two states and that each protein can interact with any other . The latter assumption allows generality in the model without imposing limitations . Every pathway structure that could be constructed in the presented model could be constructed with real biochemistry ( potentially requiring more proteins ) . Still , we have tested the effects of imposing possible limitations on the pathway structure on the evolution of taxis responses . These included imposing the requirement that the receptor can only be a kinase ( i . e . it could only activate other proteins ) , that each protein can have only a single interaction , or that self-regulation is not allowed in the model . We find such limitations not to significantly affect the outcome of the evolutionary simulations ( data not shown ) . To test the effect of having fluctuating stimulus source in the environment on the evolution of taxis responses we run additional simulations where a number of the parameters were chosen independently for each generation . These included; ( i ) a background uniform stimulus distribution , chosen from a uniform distribution [0–5 . 0] , ( ii ) the peak height of the Gaussian stimulus distribution , chosen from a uniform distribution [1 . 0–10 . 0] , ( iii ) the variance ( width squared ) of the Gaussian stimulus distribution , chosen from a uniform distribution [50 . 0–500 . 0] , and ( iv ) the initial location for the bacteria chosen at random from all points in the 100×100 space . In addition , for these runs , 20 percent of all mutations resulted in an interaction being deleted ( set equal to zero ) rather than modification through addition of a randomly chosen increment . The result of these simulations were analysed by subjecting the pathway to a specified time-course of stimulus levels , and monitoring the resulting response as shown in Figure 2 . To perform a simple estimate of the efficiency of taxis response mediated by the non-adaptive pathway dynamics ( see main text ) , we consider the equilibrium situation where there is a steady-state concentration of bacteria and stimulus at any location . At each time step a bacterium can either tumble or swim , so we can express the probability that a bacterium that starts at location will swim to location as equal to , where p1→2 is the probability that a swimming bacterium at location 1 will swim to location 2 . If we now invoke detailed balance , the overall flux of bacteria swimming from will exactly equal those swimming from , or . In general , p1→2 will be a complicated function that includes information about the previous location ( and thus the current swimming direction ) of the bacterium at , but if we ignore these correlations and assume that the bacteria are swimming in an isotropic manner in a simple space , p1→2 = p2→1 . We can further assume that for the described non-adaptive pathway dynamics , the concentration of the effector is proportional to the local stimulus level . Under these conditions , assuming the tumbling probability follows Equation 3 , it is straightforward to show that the relative concentration of bacteria at different locations is given by where C is a normalisation constant . In other words , for sufficient gain ( γm λ ) and stimulus level , the steady-state bacteria concentration is proportional to the stimulus level resulting in an efficient taxis response . | Here , we study how signalling networks mediating chemotaxis could have evolved . We simulated the evolution of virtual bacteria , which can explore their environment by alternating between swimming and tumbling . The tumbling frequency is dictated by the output of a signalling network that senses extracellular nutrient levels , while the bacteria's reproductive success is determined by their ability to find nutrients . Under conditions of abundant food , we find that bacteria quickly evolve signalling networks that enable effective chemotaxis , where increasing nutrient levels increase tumbling frequency . Our findings provide explanation for network dynamics underlying similar behaviour as observed in certain mutant strains of Escherichia coli and in other bacterial species . Conversely , wild-type E . coli respond to increasing nutrient levels by decreasing their tumbling frequency and adapting to constant attractant levels . We observe such adaptive network dynamics when we repeat evolutionary simulations under conditions of scarce food . These findings suggest that ( i ) adaptation is not necessary for effective chemotaxis , ( ii ) an ancestral minimal chemotaxis system could have used a simple coupling between the signalling network and the metabolic state , and ( iii ) environmental conditions are one of the determining factors for the evolution of adaptive responses . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"biochemistry/molecular",
"evolution",
"evolutionary",
"biology/microbial",
"evolution",
"and",
"genomics",
"microbiology/microbial",
"evolution",
"and",
"genomics",
"biochemistry/chemical",
"biology",
"of",
"the",
"cell",
"evolutionary",
"biology/animal",
"behavior",
"computational",
"biology/evolutionary",
"modeling",
"biochemistry/cell",
"signaling",
"and",
"trafficking",
"structures",
"computational",
"biology/signaling",
"networks",
"microbiology/microbial",
"physiology",
"and",
"metabolism",
"computational",
"biology/systems",
"biology"
] | 2008 | Evolution of Taxis Responses in Virtual Bacteria: Non-Adaptive Dynamics |
Chronic chagasic cardiomyopathy ( CCC ) , the main clinical sign of Chagas disease , is associated with systemic CD8+ T-cell abnormalities and CD8-enriched myocarditis occurring in an inflammatory milieu . Pentoxifylline ( PTX ) , a phosphodiesterase inhibitor , has immunoregulatory and cardioprotective properties . Here , we tested PTX effects on CD8+ T-cell abnormalities and cardiac alterations using a model of experimental Chagas’ heart disease . C57BL/6 mice chronically infected by the Colombian Trypanosoma cruzi strain and presenting signs of CCC were treated with PTX . The downmodulation of T-cell receptors on CD8+ cells induced by T . cruzi infection was rescued by PTX therapy . Also , PTX reduced the frequency of CD8+ T-cells expressing activation and migration markers in the spleen and the activation of blood vessel endothelial cells and the intensity of inflammation in the heart tissue . Although preserved interferon-gamma production systemically and in the cardiac tissue , PTX therapy reduced the number of perforin+ cells invading this tissue . PTX did not alter parasite load , but hampered the progression of heart injury , improving connexin 43 expression and decreasing fibronectin overdeposition . Further , PTX reversed electrical abnormalities as bradycardia and prolonged PR , QTc and QRS intervals in chronically infected mice . Moreover , PTX therapy improved heart remodeling since reduced left ventricular ( LV ) hypertrophy and restored the decreased LV ejection fraction . PTX therapy ameliorates critical aspects of CCC and repositioned CD8+ T-cell response towards homeostasis , reinforcing that immunological abnormalities are crucially linked , as cause or effect , to CCC . Therefore , PTX emerges as a candidate to treat the non-beneficial immune deregulation associated with chronic Chagas' heart disease and to improve prognosis .
Chagas disease ( CD ) , a neglected tropical disease caused by the protozoan parasite Trypanosoma cruzi , affects 6 to 8 million people in Latin America [1] . The cardiac form , the most frequent clinical manifestation of CD , is characterized by fibrosis with remodeling of the myocardium and vasculature , which commonly progresses to heart failure [2] . The chronic chagasic cardiomyopathy ( CCC ) is a low-grade CD8+-enriched myocarditis occurring in an inflammatory cytokine-embedded milieu [3–5] . Abnormal CD8+ T-cell function may contribute to systemic inflammatory profile and cardiac tissue lesion in the chronic phase of T . cruzi infection [6–10] . Regardless their importance for T . cruzi host resistance [11] , CD8+ T-cells gained particular attention as the major component of myocarditis in acute [12] and chronic [9 , 13] experimental T . cruzi infection and in chagasic patients with CCC [3 , 4 , 14] . Recently , we proposed that interferon-gamma ( IFNγ ) + CD8+cells exert a beneficial role , whereas perforin ( Pfn ) + CD8+ cells take part in T . cruzi-induced heart injury [9] . Additionally , we proposed that a proper therapeutic tool could interfere with distinct CD8+ T-cell populations hampering heart injury [9] . Indeed , CD8+ T-cell abnormalities and systemic inflammatory profile were reduced by administration of the anti-tumor necrosis factor ( TNF ) antibody Infliximab to a model of Chagas’ heart disease [15] . These findings unveiled that reversal of systemic immunological unbalance is a rational pathway to be explored to improve the prognosis of Chagas’ heart disease . The methylxanthine pentoxifylline ( PTX ) is a phosphodiesterase inhibitor commonly used to treat peripheral vascular diseases . PTX also shows therapeutic potential as an anti-inflammatory and anti-tumor agent [16] . PTX has previously been proposed as an adjuvant therapeutic tool for leishmaniasis , a protozoan disease with an extensive inflammatory component [17] . Further , in non-infectious heart disorders PTX has shown cardioprotective effects in association with reduced plasma levels of TNF [18 , 19] . Given the lack of an effective specific therapy , CCC is treated similarly to all other heart failure syndromes using therapies to mitigate symptoms [2] . It is proposed that CCC pathogenesis relies on a parasite-driven systemic inflammatory profile , which may reverberate in the cardiac tissue and contribute to heart dysfunction [5 , 10 , 15 , 20 , 21] . Therefore , PTX arises as a therapeutic tool to interfere with immunological unbalance and to improve the progressive functional compromise of the heart in CD . Here we tested the effects of PTX on hallmarks of immunological and heart alterations detected in CD , using a model of CCC associated with high TNF expression and CD8+ T-cell abnormalities [9 , 15 , 22] .
This study was carried out in strict accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the Brazilian National Council of Animal Experimentation ( http://www . cobea . org . br/ ) and the Federal Law 11 . 794 ( October 8 , 2008 ) . The Institutional Committee for Animal Ethics of Fiocruz ( CEUA-Fiocruz-L004/09; LW-10/14 ) approved all experimental procedures used in the present study . All presented data were obtained from three ( D2–4 ) independent experiments ( Experiment Register Books #41 and 49 , LBI/IOC-Fiocruz ) . Mice obtained from the animal facilities of the Oswaldo Cruz Foundation ( CECAL/Fiocruz , Rio de Janeiro , Brazil ) were housed under specific pathogen-free conditions in a 12-h light-dark cycle with access to food and water ad libitum . Five- to 7-week-old female C57BL/6 ( H-2b ) and C3H/He ( H-2k ) were intraperitoneally infected with 100 blood trypomastigotes ( bt ) of the Type I Colombian strain [23] of T . cruzi , and parasitemia was employed as a parameter to establish acute and chronic phases [13] . The chronically T . cruzi-infected C57BL/6 and C3H/He mice represent models of mild and severe CCC , respectively , paralleled to the degree of immunological abnormalities [22] . Sex- and age-matched noninfected ( NI ) controls were analyzed in parallel . Chronically T . cruzi-infected C57BL/6 mice showing signs of CCC were intraperitoneally injected with saline ( BioManguinhos/Fiocruz , Brazil ) containing PTX ( Trental , Sanofi-Aventis , Brazil ) ( 20 mg/kg ) or vehicle daily from 120 to 150 days postinfection ( dpi ) . For immunohistochemical staining ( IHC ) , the polyclonal antibody recognizing T . cruzi antigens and supernatants containing anti-mouse CD8a ( clone 53–6 . 7 ) and anti-mouse CD4 ( clone GK1 . 5 ) were produced in our laboratory ( LBI/IOC-Fiocruz , Rio de Janeiro , RJ , Brazil ) . Other antibodies included an anti-F4/80 polyclonal antibody ( Caltag , USA ) ; biotinylated rabbit anti-goat IgG cocktail ( KPL , USA ) ; polyclonal rabbit anti-connexin 43 ( Cx43 ) ( Sigma-Aldrich , USA ) , polyclonal rabbit anti-mouse FN ( Gibco-BRL , USA ) , biotinylated anti-mouse CD54 ( intercellular cell adhesion molecule-1 , ICAM-1 , BD Pharmingen , USA ) , biotinylated anti-rat immunoglobulin ( DAKO , Denmark ) and biotinylated anti-rabbit immunoglobulin and peroxidase-streptavidin complex ( Amersham , UK ) . Monoclonal antibodies anti-mouse Pfn ( CB5 . 4 , Alexis Biochemicals , USA ) and anti-IFNγ ( R4–6A2 , BD PharMingen , USA ) produced in rat were also used in IHS . For flow cytometry studies , PE-Cy7-anti-mouse TCRαβ ( clone H57–597 ) , APC-conjugated anti-mouse CD8a ( clone 53–6 . 7 ) , FITC-anti-CD4 ( GK1 . 5 ) , PE-rat anti-mouse TNF ( clone MP6-XT22 ) , PerCP-anti-CD4 ( clone GK1 . 5 ) , FITC- conjugated anti-Pfn ( 11B11 ) and PECy-7-conjugated anti-IFNγ ( clone XMG1 . 2 ) were purchased from BD Pharmingen ( USA ) . PE-conjugated anti-CD107a ( clone eBIO1D4B ) was obtained from eBioscience . Anti-TNF receptor ( TNFR ) 1 ( TNFR1/p55/CD120a; clone 55R-286 ) conjugated to PE was purchased from BioLegend ( USA ) . Appropriate controls were prepared by replacing the primary antibodies with the corresponding serum , purified immunoglobulin or isotype . All antibodies and reagents were used according to the manufacturers’ instructions . Spleens were minced and the red blood cells were removed using lysis buffer ( Sigma-Aldrich , USA ) . In a set of experiments , peripheral blood was also collected , as previously described [9] . The splenocytes and blood cells were labeled , events were acquired with a CyAn-ADP ( Beckman Coulter , USA ) and the data were analyzed with the Summit v . 4 . 3 Build 2445 program ( Dako , USA ) as described elsewhere [9] . The ELISpot assay for the enumeration of IFNγ-producing cells was performed in triplicate as previously described [24] . Plates were coated with anti-mouse IFNγ ( clone R4–6A2; BD PharMingen , USA ) antibody diluted in PBS ( 5 μg/mL ) . Antigen-presenting cells were primed for 30 minutes at 37°C with total frozen extracts of epimastigote forms ( Y strain ) and amastigote surface protein 2 ( ASP2 ) H-2Kb-restricted VNHRFTLV peptide [25] . After incubation , the freshly isolated splenocytes from experimental mice were seeded at 5 x 105 cells/well and incubated for 20 hours at 37°C and 5% CO2 . Biotin-conjugated anti-mouse IFNγ antibody ( clone XMG1 . 2; BD PharMingen , USA ) was used to detect the captured cytokines . Spots were revealed after incubation of the samples with a solution of alkaline phosphatase-labeled streptavidin ( BD PharMingen , USA ) and a solution of NBT and BCIP ( Sigma-Aldrich , USA ) in Tris buffer ( 0 . 9% NaCl , 1% MgCl2 , 1 . 2% Tris in H2O ) . The mean number of spots , in triplicate wells , was determined for each experimental condition . The number of specific IFNγ-secreting T-cells was calculated by estimating the stimulated spot count/106 cells using a CTL OHImmunoSpot A3 Analyzer ( USA ) . A mouse cytometric bead array ( CBA ) Inflammation Kit ( Becton & Dickinson , USA ) was used to quantify cytokines in the serum according to the manufacturer’s instructions . The fluorescence produced by the CBA beads was measured with a FACSCalibur instrument ( Becton Dickinson , USA ) and analyzed using FCAP Array software . Standard curves ( 1 pg/mL to 100 ng/mL ) were generated in parallel . This method consistently detected concentrations above 10 pg/mL . For real-time quantitative RT-PCR ( RT-qPCR ) , the hearts were harvested , washed to remove blood clots , weighed and frozen in RNAlater ( Life Technologies , USA ) . Total RNA ( for gene expression studies ) and DNA ( for parasite detection ) were extracted from the same sample using TRI-Reagent ( Sigma-Aldrich , USA ) . For detection of TNF mRNA , the reverse transcriptase reactions were performed using a SuperScript III First Strand Synthesis Kit , and RT-qPCR was performed using TaqMan gene expression assays for TNF ( # Mm00443258-m1 ) and the endogenous housekeeping control genes glyceraldehyde 3-phosphate dehydrogenase ( GAPDH ) ( # Mm99999915-g1 ) and β actin ( # Mm00607939-s1 ) , purchased from Life Technologies ( USA ) . Reactions were performed in duplicate according to manufacturer’s instruction , using cDNA template obtained from 2μg RNA . The conditions for the PCR were as follows: 95°C for 10 minutes , followed by 40 cycles at 95°C for 15 seconds and 60°C for 1 minute . Relative quantification of target gene levels was performed using the ΔΔCt method [26] . RT-qPCR data were normalized by the housekeeping genes GAPDH and β actin mRNA , using the Expression Suite Software V1 . 0 . 3 ( Life Technologies , USA ) and fold increase was determined in comparison with NI controls . For parasite detection 5 μL of purified DNA was analyzed by real time quantitative PCR ( qPCR ) using TaqMan system , with primers Cruzi 1 ( 5'-AST CGG CTG ATC GTT TTC GA-3' ) , Cruzi 2 ( 5'-AAT TCC TCC AAG CAG CGG ATA-3' ) and probe Cruzi 3: 6FAM-CACACACTGGACACCAA-MGB ) targeting the T . cruzi nuclear satellite DNA , as previously [27] . As an internal amplification control , the TaqMan assay targeting mice glyceraldehyde 3-phosphate dehydrogenase ( GAPDH ) ( # Mm99999915-g1 , Life Technologies , USA ) was used . Parasite load quantification was estimated by absolute quantification , following normalization by heart sample weight . The standard curve for the absolute quantification was generated by a 1:10 serial dilution of DNA extracted from the Colombian strain epimastigote culture stocks , ranging from 106 to 0 . 5 parasite equivalents . In a flat bottom 96-well plate ( Corning , Inc , USA ) were distributed 100 μL of RPMI containing 5 x 106 bt/well of the T . cruzi . On parasites , were added 100 μL of different concentrations of PTX ( 0 . 3 , 1 , 3 , 10 , 30 , 100 and 300 μg/mL ) , the positive ( 10 μM of the trypanocidal drug benznidazole ) and negative ( injection grade saline , Biomanguinhos/Fiocruz , Brazil ) controls . After incubation for 24 hours at 37° C in an incubator containing constant tension of 5% CO2 ( Shell Lab , USA ) , the number of parasites in the different treatment conditions was counted in a Neubauer chamber . Eight to fifteen T . cruzi-infected and three to five NI animals were euthanized under anesthesia at 120 or 150 dpi and the hearts were removed , embedded in the tissue-freezing medium Tissue-Tek ( Miles Laboratories , USA ) and stored in liquid nitrogen . The phenotypes of the inflammatory cells colonizing the heart tissue and the T . cruzi parasitism were characterized and analyzed as previously described [9] . The ICAM-1- , FN- and Cx43-positive areas in 25 fields ( 12 . 5 mm2 ) per section ( 3 sections per heart ) were evaluated with a digital morphometric apparatus . The images were digitized using a color view XS digital video camera adapted to a Zeiss microscope and analyzed with AnalySIS AUTO Software ( Soft Imaging System , USA ) . According to the analyzed parameter , the data are shown as percent of positive area in the heart , as distance ( μm ) between stained gap junctions or as numbers of parasite nests or cells per 100 microscopic fields of view ( 400 X ) . The activity of the creatine kinase cardiac MB isoenzyme ( CK-MB ) was measured using a commercial CK-MB Liquiform kit ( Labtest , Brazil ) according to the manufacturer’s recommendations as previously adapted for mouse samples [9] . Mice were tranquilized with diazepam ( 10 mg/kg ) and transducers were placed subcutaneously ( DII ) . The traces were recorded for 2 minutes using a digital Power Lab 2/20 system connected to a bio-amplifier at 2 mV for 1 second ( PanLab Instruments , Spain ) . The filters were standardized to between 0 . 1 and 100 Hz and the traces were analyzed using Scope software for Windows V3 . 6 . 10 ( PanLab Instruments , Spain ) . The ECG parameters were analyzed as previously described [9] . For analysis of cardiac function through echocardiography mice were anesthetized with 1 . 5% isoflurane gas in oxygen with flow 1L/minute , trichotomized in precordial region and examined with a Vevo 770 ( Visual Sonics , Canada ) coupled to a 30 MHz transducer . Cardiac geometry was made using two dimensional mode images acquired for measurement of internal area of heart cavities ( right and left ventricles ) . M-mode images showed left ventricular ( LV ) muscle thickness was used for measurement LV mass . Heart and LV hypertrophy were measured by the ratios of heart weight ( HW ) and LV mass to body weight ( BW ) , respectively . Left ventricular ejection fraction ( LVEF ) was determined using Simpson’s method and left and right ventricular ( LV and RV ) areas were obtained in B-mode using a short axis view at the level of the papillary muscles . Data are expressed as mean ± SD . Analysis was performed using GraphPrism ( GraphPad , USA ) . Comparison between groups was carried out by analysis of variance ( ANOVA ) followed by Bonferroni´s post-test or t-Student test when indicated . Differences were considered statistically significant when p<0 . 05 .
To determine the effect of PTX on immune response in chronic experimental Chagas’ heart disease , PTX administration to C57BL/6 mice infected with the Colombian T . cruzi strain was initiated at 120 dpi ( S1A Fig . ) . At this time-point , CD8-enriched myocarditis , splenomegaly , immune abnormalities , cytokine unbalance , electrical alterations and heart injury are already installed [9 , 15 , 22 , 28] . At 150 dpi , all PTX-treated infected mice were alive ( S1B Fig . ) . Low parasitemia was detected in chronically infected mice administered with saline or PTX ( 6 . 8 ± 2 . 2 x 104 parasites /mL in saline-injected vs 6 . 4 ± 4 . 6 x 104 parasites /mL in PTX-treated; p>0 . 05 ) . Further , PTX therapy had no effect on body weight ( 22 ± 1 . 3 g in NI controls; 21 ± 2 . 5 g in saline-injected vs 20 . 1 ± 0 . 8 g in PTX-treated infected mice; p>0 . 05 ) . In comparison with sex- and age-matched NI controls , chronically T . cruzi-infected C57BL/6 mice presented splenomegaly ( p<0 . 001 ) , which is significantly ( p<0 . 05 ) reversed by PTX therapy ( S1C Fig . ) . Chronically T . cruzi-infected C57BL/6 mice have increased levels of TNF in the serum ( S2A Fig . ) and TNF mRNA in the heart tissue ( S2B Fig . ) , corroborating previous data [15 , 22 , 28] . Further , at 150 dpi C57BL/6 mice have increased frequency of TNF-producing CD8+ T-cells in spleen ( S2C Fig . ) . One of the proposed beneficial effects of PTX is its capacity of modulate TNF production [16] . However , PTX treatment from 120 to 150 dpi had no effect on T . cruzi-induced high TNF levels in the serum ( S2A Fig . ) , TNF mRNA overexpression in the heart tissue ( S2B Fig . ) and TNF expression by CD8+ T-cells in spleen ( S2C Fig . ) . TNF signals via TNFR1/p55/CD120a and TNFR2/p75/CD120b [29] . In chronically T . cruzi-infected saline-injected C57BL/6 mice , there was a remarkable increase in the frequency of CD8+ TNFR1+ and CD8+ TNFR2+ T-cells in the spleen ( S2D Fig . ) . PTX was previously shown to downmodulate TNFR1 expression by hepatic cells [30] . Importantly , PTX therapy ( from 120 to 150 dpi ) completely abrogated the elevated frequency of TNFR1+ CD8+ T-cells . Nevertheless , PTX therapy only partially reduced the high frequency of TNFR2-bearing CD8+ T-cells detected in chronically infected mice ( S2D Fig . ) . These data support that PTX was selectively active in chronic experimental CD . At 150 dpi , although a significant reduction in splenomegaly was noticed in PTX-treated mice ( S1C Fig . ) , similar frequencies of CD8+ T-lymphocytes were detected in the spleen of NI controls and saline-injected and PTX-treated chronically infected mice ( Fig . 1A ) . Further , CD8+ T-cells also express similar density of the CD8 molecule on cell surface in the studied groups ( MFI in NI: 25 . 6 ± 0 . 6; saline-injected T . cruzi-infected: 20 . 8 ± 1 . 5; PTX-treated T . cruzi-infected: 23 . 7 ± 2 . 6 ) . However , compared with NI controls , a considerable part of the splenic CD8+ cells of chronically Colombian-infected mice expressed TCRαβLow ( Fig . 1B; p<0 . 001 ) , corroborating previous data in a distinct model of chronic CD [6] . Importantly , PTX treatment significantly rescued the downregulation of TCR expression in CD8+ T-cells of chronically T . cruzi-infected mice ( Fig . 1B ) , considering both frequency of TCRαβLow population ( p<0 . 05 ) and density of TCRαβ on cell surface ( p<0 . 05 ) . Next , we investigated whether PTX therapy influenced the expression of markers of naïve/memory/activation phenotypes of CD8+ T-cells during chronic T . cruzi infection using CD45 , a tyrosine phosphatase essential for T-cell activation , and the expression of CCR7 , a chemokine receptor associated with the homing of T-cells to lymph nodes [31] . Compared with NI controls , splenic TCR+CD8+ T-cells of chronically infected mice displayed low frequencies of CD45RA+CCR7+ ( naïve; p<0 . 01 ) cells and CD45RA-CCR7+ ( central memory; p<0 . 001 ) cells , but showed increased frequency of CD45RA-CCR7- ( effector memory; p<0 . 01 ) cells ( Fig . 1C ) . PTX treatment restored the frequencies of the minor CD45RA+CCR7+ T-cell subset ( p<0 . 05 ) , but did not interfere with the proportions of the major CD45RA-CCR7+ and CD45RA-CCR7- CD8+ T-cell populations ( Fig . 1C ) . In an attempt to further dissect the immunoregulatory mechanism of PTX in chronic experimental CD , we analyzed the frequencies of naïve , memory and activated T-cells studying the expression of CD44 , a chondroitin sulfate proteoglycan receptor associated with cell migration to peripheral tissues , and CD62L , a marker of T-cell homing to lymph nodes [31] . In chronic T . cruzi infection , there was a remarkable decrease in the frequencies of CD44-CD62L+ naïve ( p<0 . 001 ) and CD44+CD62L+ central memory ( p<0 . 01 ) but an increase in the frequency of CD44+CD62L- ( p<0 . 001 ) CD8+ T-cells ( Fig . 1D ) , in comparison with age-matched NI mice . Notably , PTX therapy partially restored these drastic alterations , significantly ( p<0 . 05 ) increased the frequencies of CD44-CD62L+ and CD44+CD62L+ cells and decreased the proportion of CD44+CD62L- CD8+ T-cells in the spleen ( Fig . 1D ) . However , there were no changes in the frequencies of CD8+ CD44+CD62L- ( 46 . 3 ± 3 . 3% in saline-injected vs 51 . 9 ± 5 . 5% in PTX-treated T . cruzi-infected mice; p>0 . 05 ) and CD8+ CD44+CD62L+ ( 15 . 4 ± 3 . 3% in saline-injected vs 10 . 7 ± 3 . 4% in PTX-treated T . cruzi-infected mice; p>0 . 05 ) T-cells in the blood of C57BL/6 mice . Next , we explored the potential influence of PTX on the effector function of T . cruzi-specific total and CD8+ T-cells of chronically T . cruzi-infected C57BL/6 mice . Using ELISpot assay , we detected an increased ( p<0 . 01 ) number of T-cells producing IFNγ after recognition of crude T . cruzi antigens ( epimastigote extracts ) in chronically infected mice . Also , the number of IFNγ-producing CD8+ T-cells specific for the immunodominant H-2Kb-restricted ASP2 VNHRFTLV peptide was increased ( p<0 . 001 ) ( Fig . 2A ) . PTX therapy did not alter the number of IFNγ-producing cells among splenocytes recognizing crude T . cruzi antigens , but upregulated ( p<0 . 05 ) the number of IFNγ-producing ASP2-specific CD8+ T-cells ( Fig . 2A ) . Having in mind the different CD8+ T-cell phenotypes , we evaluated the potential cytotoxic activity by CD8+ splenocytes of chronically infected mice , studying the expression of CD107a , a marker for T-cell degranulation [32] , and inflammatory potential , studying intracellular IFNγ expression . At 150 dpi , in comparison with NI controls , there was a significant increase in the frequencies of IFNγ+ ( p<0 . 01 ) and IFNγ+CD107a+ and CD107a+ ( p<0 . 05 ) CD8+ T-cells in saline-injected infected mice ( Fig . 2B and S3 Fig . ) . PTX therapy significantly increased the frequency of IFNγ+ CD8+ T-cells ( p<0 . 01 ) and reduced the frequency of CD107a+ ( p<0 . 01 ) CD8+T-cells ( Fig . 2B and S3 Fig . ) . Considering the antagonistic roles for IFNγ+ and Pfn+ CD8+ T-cells in T . cruzi infection [9] , we studied the effect of PTX on IFNγ and Pfn expression by CD8+ T-cells . PTX administration to chronically infected mice increased ( p<0 . 05 ) the frequency of IFNγ+ cells , but reduced ( p<0 . 05 ) the frequencies of IFNγ+Pfn+ and Pfn+ CD8+ T-cells ( Fig . 2C ) . To investigate whether PTX effects on inflammatory IFNγ and cytotoxic Pfn+ cells were restricted to splenic compartment , we analyzed the numbers of IFNγ+ and Pfn+ inflammatory cells invading the heart tissue of chronically infected C57BL/6 mice . Saline-injected and PTX-treated chronically infected mice had similar numbers of IFNγ+ cells in the heart tissue ( Fig . 2D ) . In comparison with saline injection , PTX therapy reduced ( p<0 . 05 ) the number of inflammatory Pfn+ cells infiltrating the cardiac tissue of chronically infected mice ( Fig . 2D ) . Our previous data support that the formation of CD8-enriched chagasic myocarditis involves CCR1/CCR5-mediated cell migration [33 , 34] . Further , the CCR5 receptor and the cell adhesion molecule LFA-1 are co-expressed by peripheral blood mononuclear cells enabling them to migrate to heart tissue [33 , 35] . Here we described an increased ( p<0 . 001 ) frequency of CD8+ T-cells co-expressing LFA-1 and CCR5 among CD8+ T-cells in chronically infected mice . Importantly , PTX treatment led to a significant reduction in the frequency of splenic LFA-1+CCR5+ CD8+ T-cells ( Fig . 3A ) , when compared with saline injection . Moreover , a significant reduction in the frequency LFA-1+CCR5+ was also observed among circulating CD8+ T-cells after PTX therapy ( 3 . 14 ± 1 . 0% in PTX-treated vs 5 . 75 ± 2 . 2% in saline-injected T . cruzi-infected mice; p<0 . 05 ) . Next , we analyzed the expression of ICAM-1 , the LFA-1 ligand , on the cardiac endothelial cells [36] . ICAM-1 is upregulated ( p<0 . 001 ) in the endothelial cells of heart blood vessels and cardiomyocytes of chronically T . cruzi-infected C57BL/6 mice ( Fig . 3B and Fig . 3C ) . After PTX therapy , reduction ( p<0 . 01 ) in ICAM-1 expression was noticed in cardiomyocytes and inflammatory cells infiltrating the heart tissue ( Fig . 3C—upper panel ) and , particularly , in the blood vessel endothelial cells ( Fig . 3C—bottom panel ) . ICAM-1+ blood vessels with perivascular cuffs with several layers of inflammatory cells were commonly found in the heart of saline-injected but absent in PTX-treated T . cruzi-infected mice ( Fig . 3B and Fig . 3C ) . At 150 dpi , the Colombian-infected C57BL/6 mice present myocarditis ( Fig . 3D ) , mainly composed of CD8+ T-cells , corroborating previous findings [9] . Interestingly , the short term PTX therapy reduced ( p<0 . 05 ) the intensity of the chronic T . cruzi-induced myocarditis ( Fig . 3D ) . The pathogenesis of Chagas’ heart disease is , at least in part , accounted to parasite persistence [2 , 37] . To bring further mechanistic insights into the beneficial effects of PTX in chronic infection , we analyzed a putative effect of PTX directly on T . cruzi trypomastigote forms and on heart parasitism . Contrasting with a significant ( p<0 . 001 ) effect of the trypanocidal drug Bz ( positive control ) , PTX showed no direct action on the survival of the trypomastigote forms of the parasite , in an in vitro assay ( S4A Fig . ) . Meanwhile , PTX did not interfere with parasite control in chronically infected C57BL/6 mice , as rare T . cruzi amastigote nests ( S4B Fig . ) and low numbers of parasite DNA copies ( S4C Fig . ) were similarly ( p>0 . 05 ) detected in the heart tissue of saline-injected and PTX-treated mice , at 150 dpi . All the beneficial effects of PTX on the unbalanced immune response of chronically infected mice encouraged us to investigate the effects of PTX therapy on T . cruzi-induced chronic heart injuries . Chronically T . cruzi-infected mice showed a significant reduction ( p<0 . 01 ) in Cx43 expression in the intercalary disc of myocardial cells , seen as increased distance of the Cx43-bearing gap junction plaques ( Fig . 4A and Fig . 4B ) . Further , chronically infected mice presented FN overdeposition in the heart tissue ( p<0 . 001 , Fig . 4A and Fig . 4C ) , and CK-MB activity levels in the serum ( p<0 . 05 , Fig . 4D ) , compared with age- and sex-matched NI controls . In comparison with saline injection , PTX therapy ameliorated heart tissue injuries , improving Cx43 expression ( p<0 . 05 , Fig . 4A and Fig . 4B ) , decreasing FN overexpression ( p<0 . 01 , Fig . 4A and Fig . 4C ) and reducing CK-M activity in the serum ( p<0 . 01 , Fig . 4D ) . Thus , PTX therapy hampered the progression of heart injury in chronically T . cruzi-infected C57BL/6 mice . Moreover , considering that significant increase in CK-MB activity is already detected in T . cruzi-infected C57BL/6 mice at 120 dpi [9] , our data support that PTX therapy reversed cardiomyocyte injury . After the demonstration that PTX restored major immunological abnormalities believed to be associated with the severity of Chagas’ heart disease [9 , 10 , 15 , 21 , 22] , we examined the effect of PTX on electrical conduction in an experimental model of CCC . When compared with sex- and age-matched NI controls , saline-injected chronically infected C57BL/6 mice presented ECG alterations including prolonged P wave , PR interval and QRST complex ( Fig . 5A ) . At 150 dpi , PTX-treated mice improved ECG alterations , compared with saline-injected mice ( Fig . 5A ) . Notably , PTX had beneficial effects on heart rate ( p<0 . 05 ) , PR ( p<0 . 001 ) and QRS ( p<0 . 05 ) intervals in comparison with saline-injected animals ( Fig . 5B ) . Actually , PTX therapy reduced the proportion of mice afflicted by arrhythmias ( ART ) , second-degree atrio-ventricular block ( AVB2 ) and other ECG abnormalities ( Fig . 5C ) . At 120 dpi , ECG abnormalities are already detected in T . cruzi-infected C57BL/6 mice [9]; hence , our data support that PTX therapy reversed ECG alterations . In parallel experiments , parasitemia , heart parasitism and inflammation were higher in Colombian-infected C3H/He compared with C57BL/6 mice , at 120 dpi [22] . In the model of severe infection ( C3H/He ) , PTX therapy also ameliorated the expression of the biomarkers of heart injury Cx43 loss and FN deposition in the heart tissue , CK-MB activity levels in the serum ( p<0 . 01 , S5A Fig . and S5B Fig . ) and ECG abnormalities ( p<0 . 05 , S5C Fig . and S5D Fig . ) . At 150dpi , the increased P duration , QTc and QRS intervals ( p<0 . 05 , S5C Fig . ) and the proportions of mice afflicted by ECG abnormalities were diminished by PTX therapy ( S5D Fig . ) . At 150 dpi , all analyzed groups of C57BL/6 mice showed similar body weight ( NI: 23 . 2 ± 1 . 3; saline-injected T . cruzi-infected: 20 . 6 ± 1 . 3; PTX-treated T . cruzi-infected: 22 . 5 ± 2 . 0 ) . Chronic T . cruzi infection resulted in heart enlargement , shown as increased HW/BW ratio ( Fig . 6A ) , corroborating our previous data [9] . The HW/BW coefficient of PTX-treated mice tends to decrease when compared with saline-injected infected mice and was similar to the HW/BW coefficient of sex- and age-matched NI controls ( p>0 . 05; Fig . 6A ) . To assess whether chronic T . cruzi infection affects heart geometry and function , as well as the impact of PTX administration , all chronically infected mice underwent echocardiographic evaluation . At 150 dpi , compared with NI C57BL/6 controls , infected mice showed alterations in heart geometry , as increased LV mass ( NI: 82 . 4 ± 2 . 9; saline-injected T . cruzi-infected: 100 . 7 ± 4 . 3; p<0 . 01 ) . PTX therapy significantly reduced LV mass ( PTX-treated T . cruzi-infected: 80 . 1 ± 4 . 3; p<0 . 01 ) , restoring mass values akin to NI controls . The assessment of heart hypertrophy , as the ratio of LV mass to body weight , showed a remarkable LV hypertrophy ( p<0 . 01 ) in chronically Colombian-infected mice ( Fig . 6B ) . Importantly , PTX therapy significantly reduced the LV hypertrophy ( p<0 . 01 ) of T . cruzi-infected mice ( Fig . 6B ) . Additionally , compared with NI controls , saline-injected chronically infected mice showed higher right ventricular ( RV ) and LV areas ( Fig . 6C ) . PTX-treated mice exhibited RV and LV areas similar to age-matched NI controls ( Fig . 6C ) . When compared with NI age-matched controls , mice chronically infected by the Colombian strain showed significant decrease in LVEF ( Fig . 6D ) . Notably , PTX therapy significantly ( p<0 . 05 ) restored LVEF of chronically infected mice to values resembling NI controls ( Fig . 6D ) .
Here we used PTX as a therapeutic tool in experimental CCC , exploring the effects of this immunomodulator upon key features of immunological and heart abnormalities . After PTX therapy , the overexpression of TNF remained unaltered , while TNFR1 and TNFR2 expression was reduced . PTX therapy also decreased the frequency of splenic cytotoxic ( CD107a+ and Pfn+ ) CD8+ T-cells and the number of Pfn+ cells invading the cardiac tissue . Conversely , treatment with PTX increased the number of IFNγ-producing ASP2-specific CD8+ T-cells . Further , PTX-treated mice show reduced frequency of splenic and circulating LFA-1+CCR5+ CD8+ T-cells , decreased expression of ICAM-1 on cardiac tissue and less severe myocarditis . Importantly , PTX therapy ameliorated heart injury and dysfunction , without interfering with parasite control . Splenomegaly is present in chronic CD patients [38] . In experimental T . cruzi infection , splenomegaly is associated with increased polyclonal T-cell activation , a hallmark of not regulated immune response [6 , 9 , 38] . Notably , the splenomegaly seen in Colombian-infected C57BL/6 mice [9 , 15] was partially reversed by PTX therapy . This effect may result of the inhibitory effect of PTX on primary proliferative capacity of T-cells [39] . PTX ameliorates LVEF in patients with noninfectious heart failure in association with downmodulation of inflammatory biomarkers , such as TNF [18 , 19] . Elevated plasma TNF levels paralleled CCC severity in patients [21 , 40] and chronic experimental CD models [15 , 22] . During acute T . cruzi infection , PTX administration reduced the number of TNF+ cells in necrotic areas in the spleen [41] . Thus , a reduction in TNF expression was expected in our models of CCC subjected to PTX treatment . However , this was not the case as PTX did not affect TNF overexpression in the heart tissue or systemically ( spleen and serum ) . The lack of effect of PTX on TNF expression in experimental CCC was not a surprise . Clinical trials assessing PTX in chronic heart failure showed no concordant action on downmodulation of TNF levels , despite clinical improvement and beneficial effects on biomarkers of heart lesion [16] . The mechanism by which PTX improves heart diseases remains unsolved [16 , 18 , 19] . TNF signals through two different receptors ( TNFR1 and TNFR2 ) and affects diverse biological processes such as cell activation , proliferation , differentiation and survival [29] . Further , depending on the biological process TNF receptors may have opposite effects [42] . In CCC , PTX abrogated TNFR1 expression by CD8+ T-cells . Also , in ischemic-reperfusion hepatic injury the beneficial effect of PTX was associated with abrogation of the high expression of TNFR1 mRNA [30] . Thus , the downmodulatory effect of PTX on TNFR1 occurs independently of the trigger and target cell type . Additionally , in chronically infected mice PTX therapy partially reduced the upregulated expression of TNFR2 by CD8+ T-cells . Since TNFR2 receptor is involved in cell proliferation [29] and upregulated after T-cell activation [43] , in reducing the frequency of TNFR2-expressing cells may reside , at least in part , the beneficial effect of PTX on splenomegaly during chronic T . cruzi infection . Therefore , akin the effect of anti-TNF therapy in CCC [15] , PTX may emerge as a tool to disrupt TNF/TNFRs signaling pathway and restore immunological homeostasis and cardiac injury in experimental CCC . CD8+ T-cells are the prominent inflammatory components in the cardiac tissue in Chagas’ heart disease [3 , 4 , 14] , a feature reproduced in Colombian-infected mice [9 , 13 , 15 , 28 , 34] . In CD patients , CD8+ T-cells show abnormal activation phenotypes marked by low expression of CD8 and TCR [7] . Although in chronically Colombian-infected mice no changes in the frequency of splenic CD8+ T-cells and in the density of CD8 molecules on cell surface were detected , downmodulation of TCRαβ expression on CD8+ T-cells was noticed , corroborating previous data [6] . Importantly , in PTX-treated mice the frequency of TCR-bearing cells and the density of TCRαβ on cell membrane were restored to intensities similar to NI controls . In naïve T-cells , TCR complex is constitutively internalized and rapidly recirculate back to cell surface . However , by a molecular process not yet fully understood , antigen stimulation increases retention/degradation of TCR lowering the density of TCR on cell surface in association with reduced effector function [44 , 45] . Therefore , in chronic T . cruzi-infection PTX therapy mice may interfere with the turnover of TCR and restore the capacity of CD8+ T-cells to respond to activation signals . This idea deserves to be explored . In comparison with noninfected individuals , CD patients show significant increase in total effector/memory CD8+ T-cells ( CD45RA−CCR7− ) , supporting a continuous stimulus by parasite antigens [7] . Similarly , increased frequency of effector/memory CD45RA−CCR7− CD8+ T-cells was detected in chronically infected C57BL/6 mice . PTX did not interfere with the frequency of these cells , supporting that they are potentially prone to control invasive pathogens [31] . Conversely , PTX therapy restored the naive CD45RA+CCR7+ CD8+ T-cell compartment , potentially seeking homeostasis and allowing immune response to new stimuli . Here we corroborated the findings that chronic T . cruzi infection reduced the frequency of CD62L+ ( naïve/central memory ) and increased the percentage of CD44+CD62L− ( activated ) cells among CD8+ T-cells [13 , 46] . Interestingly , PTX therapy decreased the frequency of CD44+CD62L− and increased the proportion of CD44−CD62L+CD8+ splenocytes . Therefore , PTX interferes with cell migration and increases retention of T-cells in secondary lymphoid tissues , major sites of antigen recognition [47] . In chronic T . cruzi infection , another remarkable disturb in cell migration scenario is the increased frequency of splenic and circulating LFA-1+CCR5+ CD8+T-cells [9 , 13 , 34] . These cells are potentially able to invade the heart tissue , where CCR5 ligands ( CCL3/MIP-1α and CCL5/RANTES ) are found [13 , 28 , 34] . As PTX therapy lowered the frequencies of splenic and circulating LFA-1+CCR5+ CD8+ T-cells in chronically infected mice , a reduction in myocarditis intensity was expected . Indeed , PTX decreased the number of inflammatory cells invading the heart tissue . T . cruzi infection increases the expression of ICAM-1 ( ligand of LFA-1 ) on cardiac endothelial cells and cardiomyocytes [13 , 48] , aiding T-cell entry into the heart tissue [35] . In chronically infected mice , PTX decreased the expression of ICAM-1on cardiac tissue . PTX was also shown to downmodulate ICAM-1 in a model of acute lung injury [49] . Therefore , in chronically infected mice PTX may reduce the CCR5-mediated chemotaxis and LFA-1/ICAM-1-mediated endothelial cell/lymphocyte interaction , explaining the decreased colonization of the heart tissue by inflammatory cells . CCR5 expression is related to CCC in patients [50] and myocarditis intensity and heart injury in infected mice [33 , 34] . In T . cruzi infection the majority of the detrimental Pfn+CD8+ cells are LFA-1+CCR5+ [9] , hence we predicted a beneficial effect of PTX in experimental CCC in association with a selective reduction of the migration of Pfn+ cells to heart tissue . In experimental T . cruzi infection , antigen specific Pfn+CD8+ T-cells may play a non-beneficial role , whereas IFNγ+CD8+ T-cells may exert a protective role in heart injury [9] . In CD patients , the frequency of IFNγ-producing CD8+ T-cells specific for T . cruzi antigens is inversely related with disease severity [7] . Therefore , we tested the impact of PTX on specific T-cell response in chronically infected mice . Notably , PTX therapy increased the number of IFNγ-producing anti-T . cruzi VNHRFTLV ASP2 effector CD8+ T-cells , a population shown to protect against T . cruzi infection [51] . Previously , PTX was shown to increase T-cell memory and protective immunity against Salmonella infection [52] . In refractory patients , PTX combined with antimonial drug ameliorated cutaneous leishmaniasis [17] , supporting the use of PTX as immunological adjuvant . In chronically T . cruzi-infected mice , PTX therapy also increased the frequency of splenic IFNγ-producing CD8+ T-cells and reduced the proportion of CD8+ T-cells expressing CD107a , a marker for T-cell degranulation and cytotoxic activity [32] . Moreover , PTX treatment decreased the frequencies of Pfn+ and multifunctional Pfn+IFNγ+ CD8+ T-cells . Based on the possible antagonistic role of CD8+T-cells expressing IFNγ and Pfn , [9 , 15] , we then analyzed the influence of PTX on the composition of chronic heart inflammation . PTX-treated mice had reduced number of Pfn+ cells , although the number of IFNγ+ cells remained unaltered . The heart infiltrating Pfn+ cells , probably CD8+ T-cells acting as CTLs , are involved in tissue damage in experimental CD [9 , 15] . Interestingly , in CD patients with severe cardiomyopathy the presence of cells expressing granzyme A ( another component of the lytic machinery of CTL CD8+ T-cells ) in heart lesions is in accordance with concepts that involve cytolysis in pathogenesis of CCC [9 , 53] . Further , there is a good correlation between the numbers of IFNγ+ and CD8+ cells infiltrating the heart tissue in CD patients presenting successful parasite control [53] . Actually , PTX did not influence cardiac or systemic parasite load , reinforcing that an effective immune response , which contributes to T . cruzi control , is preserved and disconnected from factors causing heart injury [9 , 15 , 33 , 34] . Altogether , these findings support that PTX therapy in experimental CCC reduced the ability of CD8+ T-cells to migrate and invade the heart tissue , which is less permissive to lymphocyte interaction . Further , PTX diminished the frequency of activated but increased the frequency of naïve CD8+ T-cells , which is paralleled by regain of TCR density on CD8+ lymphocytes and , apparently , restored the capacity to respond to new antigenic stimuli . Indeed , PTX therapy increased the number of CD8+IFNγ+ responsive to ASP2 T . cruzi antigen , while disfavored Pfn+ cells inside the heart tissue . Thus , PTX reestablished several aspects of the CD8+ T-cell alterations in chronically T . cruzi-infected mice . Considering that immunological abnormalities may contribute to cardiac alterations in experimental CCC [15 , 54] , we expected that the PTX-induced amelioration of the immunological unbalance in chronic T . cruzi infection would beneficially reverberate in the cardiac injury and dysfunction . A loss of Cx43 , the most abundant ventricular gap junction protein , is associated with arrhythmogenic disease [55] . The Cx43 loss may contribute to electrical conduction abnormalities in Chagas’ heart disease [56] . One important beneficial effect of PTX was the restoration of gap-junction Cx43 expression in chronically infected mice , therefore , indicating that Cx43 loss may be interrupted . In CD , overdeposition of FN discloses cardiac fibrosis [57] . PTX therapy hampered the progression of FN overexpression in experimental CCC; reinforcing the idea that in T . cruzi infection cardiac fibrosis can be improved , and even , reversed [15 , 54 , 58 , 59] . The increased CK-MB activity in the serum , an important CCC feature and a biomarker of cardiomyocyte lesions [60] , is increased in chronically infected mice before therapy , at 120 dpi [9] . PTX therapy also reduced and , moreover , reversed the increased CK-MB activity in the serum of experimental CCC . These findings support a broad beneficial effect of PTX on major features of T . cruzi-triggered heart injury . To our knowledge , this study is the first demonstration that PTX improves electrical conduction and heart dysfunction in an infectious cardiomyopathy . In chronically infected C57BL/6 mice showing signs of CCC [9] , PTX therapy ameliorated bradycardia , prolonged P wave duration , PR and QRS intervals , ART and AVB2 . As previous shown in patients [2 , 61] and T . cruzi-infected rhesus monkeys [59] , higher ECG QRS scores directly correlated with the severity of heart fibrosis . Considering that electrical abnormalities and FN overdeposition were detected at 120 dpi [9 , 54] , when therapy was initiated , our data support that PTX more than hampering progression is reversing electrical abnormalities and heart injury in experimental CCC . Moreover , the beneficial effects of PTX therapy were not restricted to a particularly experimental model , as amelioration of heart injury and electrical alterations were also observed in Colombian-infected C3H/He mice , a model of severe CCC [22] . Lastly , ECO studies revealed anatomical alterations with increased RV and LV areas , higher LV mass and decreased LVEF in chronically Colombian-infected mice , when compared with age-matched NI controls . Dilatation of the RV is a risk factor for sudden death in several cardiac diseases [62] . The enlargement of the RV , a marker for CCC in mice [63 , 64] , was reduced by PTX therapy in chronic experimental CD . Increased LV internal dimensions emerged as a risk factor associated with morbidity and mortality in CD [65] . Further , increased LV mass , a marker of hypertrophy , is an independent risk factor of cardiovascular events [66] . Remarkably , PTX administration to mice with signs of CCC ameliorated the alterations in LV geometry and mass . Moreover , the reduced LVEF seen in chronically infected mice was improved to values detected in age-matched NI controls after PTX therapy . Similarly , PTX therapy improved LVEF in patients with heart failure due to ischemic cardiomyopathy [18 , 19] . Altogether , these data support that PTX therapy in chronic experimental CCC bettered cardiac tissue injury , electrical abnormalities and heart failure . Here we bring evidence that immunological unbalance and Chagas’ heart disease are interconnected , involving multifactorial elements that may be working bidirectionally [9 , 15 , 22 , 33] . Further , whether the beneficial effects of PTX results of its immunomodulatory properties or direct action on heart tissue , particularly on cardiomyocytes , remains to be clarified . Therefore , our results opened a new avenue to be paved to explore PTX as adjuvant to immune protective response or cardioprotective tool in an infectious cardiomyopathy . More important , PTX emerges as a potent adjuvant to treat heart failure in CD . PTX might be a non fantasious strategy for CD immunotherapy , combined or not with trypanocidal drug , hampering the deleterious inflammation but preserving the beneficial anti-parasite immunity . | Chronic chagasic cardiomyopathy ( CCC ) is the main clinical manifestation of Chagas disease ( CD ) , a neglected illness caused by the protozoan parasite Trypanosoma cruzi . More than hundred years after its discovery , CD continues to be a public health problem and millions of chronically infected people wait for an effective treatment . Chagasic cardiomyopathy is associated with CD8+ T-cell-enriched myocarditis , fibrosis and cardiac electrical and structural abnormalities , frequently progressing to heart failure . Presently , the available therapies only mitigate symptoms of CCC . Abnormalities in CD8+ T-cell compartment are present in CCC patients . Recently , we described the importance of CD8+ T-cells in the pathogenesis of CCC . Therefore , our proposal was to interfere with abnormalities of CD8+ T-cells glimpsing a better prognosis for CCC . Using PTX , an affordable drug with immunomodulatory properties on T-cells and cardioprotective effects in non-infections disease , we bring a therapeutic candidate for treating CCC . PTX therapy downmodulated detrimental CD8+ T-cells and promoted T . cruzi-specific interferon-gamma-producing T-cells . Importantly , chronic chagasic electrical and echocardiographic alterations were reversed by PTX therapy . Future studies may test the use of PTX combined with trypanocidal drug or as a vaccine adjuvant to improve the quality of life of chronic CD patients . | [
"Abstract",
"Introduction",
"Material",
"and",
"Methods",
"Results",
"Discussion"
] | [] | 2015 | Pentoxifylline Reverses Chronic Experimental Chagasic Cardiomyopathy in Association with Repositioning of Abnormal CD8+ T-Cell Response |
The HMG-box factor Tcf1 is required during T-cell development in the thymus and mediates the nuclear response to Wnt signals . Tcf1−/− mice have previously been characterized and show developmental blocks at the CD4−CD8− double negative ( DN ) to CD4+CD8+ double positive transition . Due to the blocks in T-cell development , Tcf1−/− mice normally have a very small thymus . Unexpectedly , a large proportion of Tcf1−/− mice spontaneously develop thymic lymphomas with 50% of mice developing a thymic lymphoma/leukemia at the age of 16 wk . These lymphomas are clonal , highly metastatic , and paradoxically show high Wnt signaling when crossed with Wnt reporter mice and have high expression of Wnt target genes Lef1 and Axin2 . In wild-type thymocytes , Tcf1 is higher expressed than Lef1 , with a predominance of Wnt inhibitory isoforms . Loss of Tcf1 as repressor of Lef1 leads to high Wnt activity and is the initiating event in lymphoma development , which is exacerbated by activating Notch1 mutations . Thus , Notch1 and loss of Tcf1 functionally act as collaborating oncogenic events . Tcf1 deficiency predisposes to the development of thymic lymphomas by ectopic up-regulation of Lef1 due to lack of Tcf1 repressive isoforms and frequently by cooperating activating mutations in Notch1 . Tcf1 therefore functions as a T-cell–specific tumor suppressor gene , besides its established role as a Wnt responsive transcription factor . Thus , Tcf1 acts as a molecular switch between proliferative and repressive signals during T-lymphocyte development in the thymus .
Cancers often develop as consequence of deregulated expression of key factors that operate during normal development . Deregulation of the Wnt signaling pathway has been implicated in many types of malignancies , especially in solid tumors ( reviewed in [1]–[3] ) . Mutations in different components of the Wnt pathway are found to contribute to carcinogenesis [3] . During normal development , Wnt proteins function as proliferation-inducing growth factors and may also affect cell-fate decisions [4]–[6] . Wnt proteins bind to their Frizzled receptors , thereby preventing proteosomal degradation of the Wnt mediator β-catenin . Subsequently , β-catenin is translocated to the nucleus , where it forms an active transcription complex with the nuclear proteins downstream of the Wnt pathway: TCF1 ( T-cell factor 1 , the product of the Tcf7 gene , referred to as Tcf1 throughout this article ) , LEF1 ( Lymphocyte-Enhancer-binding Factor ) , or the homologous factors TCF4 and TCF3 . All TCF/LEF factors belong to a family of high-mobility-group ( HMG ) proteins that utilize the HMG box for sequence-specific DNA binding . The HMG boxes of these factors are virtually identical and likely display indistinguishable DNA-binding specificities [7] , [8] . TCF/LEF nuclear proteins exist as transcriptional repressors and only upon binding to β-catenin will form an active transcription complex . For TCF1 , at least eight isoforms have been identified with different capacities to bind β-catenin , thereby influencing the responsiveness of cells toward Wnt signals [9] . The long isoforms of TCF1 contain the amino-terminal β-catenin-binding domain , whereas the shorter isoforms lack this domain and will therefore function as the naturally occurring repressors of the pathway . A large body of evidence has shown that canonical Wnt signaling is essential for thymocyte proliferation and normal T-cell development [10]–[16] . Among the Wnt proteins , specifically Wnt1 and Wnt4 are essential for thymocyte proliferation [11] , which is reflected in mice deficient for Wnt1 and Wnt4 that display low thymic cellularity [15] . In addition , overexpression of Wnt4 selectively expands thymic output from transduced hematopoietic stem cells [17] . Recently , we showed that another Wnt protein , Wnt3a , plays a crucial role in fetal thymopoiesis , with Wnt3a−/− thymi showing severely reduced numbers of DP and a block of the preceding CD8+ Immature Single Positive ( ISP ) stage [18] , thereby displaying an exact phenocopy of fetal thymi in Tcf1−/− mice . Studies on mice deficient for the Wnt-responsive nuclear proteins reveal crucial roles for Tcf1 in T-cell development and Lef1 in B-cell development [19] , [20] . Tcf1−/− mutant mice have a severe reduction of thymic cellularity and a partial block in thymocyte differentiation at the transition from the CD8+ ISP stage to the CD4+CD8+ double positive ( DP ) stage [19] . Thymocytes of Tcf1−/− mice do not proliferate as strong as their wild-type counterparts [21] . These data indicate that lack of Tcf1 mainly results in lack of proliferation and therefore expansion of the thymocytes . Although Lef1−/− mice have normal T-cell development , mice deficient in both Lef1 and Tcf1 have a complete block in T-cell differentiation at the ISP stage , which indicates redundancy between these two factors [22] . The block in T-cell development in Tcf1-deficient mice was shown to be caused by lack of Wnt mediated signals , as Tcf isoforms without the β-catenin binding domain could not restore T-cell development , but Tcf isoforms containing the interaction domain with the Wnt mediator β-catenin fully reconstituted T-cell development [23] . In addition , soluble Frizzled receptors acting as inhibitors of Wnt signaling [11] , or overexpression of an inhibitor of the interaction between β-catenin and Tcf/Lef factors , ICAT [13] , inhibited T-cell development at the same stages as Tcf1 KO mice . Recent studies by the laboratories of Bhandoola and Gounari further emphasize the importance of Tcf1 as a critical regulator of T-lineage specification and differentiation . These investigators demonstrate that Tcf1 is critical for induction of a T-cell-specific gene program in stem cells and uncommitted progenitors [16] . In addition , the Gounari lab showed that ETPs lacking Tcf1 fail to develop normally [10] . Together , these studies conclusively point to Tcf1 as an essential transcriptional regulator of T-cell specification , commitment , and lineage determination [24] . Here we report that Tcf1 , besides acting as a Wnt responsive transcription factor , also has an important other function , namely as tumor suppressor for the development of T-cell lymphomas .
The generation of mice lacking the Wnt-responsive factor Tcf1 revealed a crucial role for Tcf1 in T-cell development [19] . Tcf1−/− mice have thymi characterized by low cellularity ( fewer than 107 cells at 6–8 wk of age , compared to >108 cells in littermates [19] ) , which is due to the blocks at the DN and ISP developmental stages . Strikingly , over time an increasing number of Tcf1−/− mice were found with an extremely enlarged thymus ( example in Figure 1A ) . The occurrence of these enlarged thymi in Tcf1−/− mice was not a rare finding . Studying the thymi of 150 Tcf1−/− mice showed a clear bimodal distribution in thymic cellularity ( Figure 1B ) . A threshold in thymocyte numbers occurs at 18×106 thymocytes . A cellularity of <18×106 cells can therefore be regarded as a normal size Tcf1−/− thymus , whilst a thymus with a cellularity >18×106 cells can be regarded as an abnormal enlarged thymus . The right graph of Figure 1B demonstrates an increasing percentage of Tcf1−/− mice with a hyperplastic thymus with increasing age . This hyperproliferation could be caused by increased normal proliferation or by the presence of a clonal population of tumor cells . Results described below collectively demonstrate that these cells are neoplastic in nature and represent thymic lymphomas ( Figures 1C and 2 ) . Immunohistochemical analysis shows that neoplastic cells completely disrupted the thymic architecture ( Figure 1C ) , and loss of corticomedullary demarcation was evident . Neoplastic cells invaded the thymic capsule , neighboring adipose tissue , thoracic organs , liver , kidney , spleen , and lymph nodes . Abdominal organs ( liver and kidney ) and lymphatic tissues ( spleen and lymph nodes ) have been shown to be preferential sites for metastasis of systemic lymphomas [25] . The lymphomas have different phenotypic characteristics that to some extend reflect the developmental blocks . Thus , the different lymphomas in the Tcf1−/− thymi were categorized into several phenotypically distinct subgroups or mixtures thereof: DN1 , DN3 , ISP , or the DP stage ( examples of DN3 , ISP , and DP lymphomas are shown in Figure 2A ) . The different phenotypes are not correlated to the age of the mice , and their frequency is: 5% DN1 , 32 . 5% DN3 , 40% ISP , and 22 . 5% DP ( n = 40 Tcf1−/− tumor mice ) . The thymocytes overpopulating the thymus were present in other hematopoietic organs such as spleen , bone marrow , and lymph nodes ( Figure 2A ) , suggesting the high malignant capacity of these cells to invade other organs as expected from the size of the organs . To examine whether this aggressive proliferation of thymocytes was due to clonal expansion , a Southern Blot analysis was performed , using the Jβ2 region of the TCRβ gene . In contrast to DNA from a Tcf1+/− control thymus , which shows the germline band ( g . l . indicated by the arrow ) and a heterogeneous mix of bands characteristic of a polyclonal cell population ( Figure 2B , left panel , Lane 1 ) , the lymphoma samples only showed between two and four distinct bands ( germline and one or more rearranged alleles ) , indicating that they consisted of one or two independent clones ( Figure 2B , left panel , Lanes 2–5 ) . Interestingly , the same clonal band was found in the metastases in the secondary organs , BM , spleen , liver , and lymph nodes ( Figure 2B , right panel ) . To further confirm the malignancy of these thymic lymphomas and their ability to grow autonomously and invade the organs in secondary recipients , 5×105 thymocytes were transferred into sublethally irradiated Rag1−/− recipients . The malignant donor thymocytes were derived from Tcf1−/− mice , characterized by intermediate expression of CD3 and CD25 , and control donor thymocytes were obtained from Tcf1+/− mice . Four weeks after transfer , the tumor cells ( as characterized by the expression of CD3 and CD25 ) were present in peripheral blood in 50% of the recipient mice ( Figure 2C ) . Six weeks after transfer , all animals were sacrificed . Recipients receiving the malignant thymocytes of Tcf1−/− origin all displayed an enlarged liver and spleen , and tumor cells were detectable by flow cytometry in all organs tested ( thymus , BM , spleen; Figure 2D ) . Together these results demonstrate that a lack of Tcf1 predisposes mice to a high risk of developing thymic lymphomas , which are clonal and characterized by an aggressive metastatic phenotype . These results indicate that Tcf1 functions as a tumor suppressor gene in the thymus . To gain insight into the molecular mechanism underlying the Tcf1-deficient tumor development , we compared the gene expression profile of thymocytes derived from Tcf1−/− mice with tumors , Tcf1−/− mice of similar age without tumors , and control Tcf1+/− mice . Samples of 17 mice were studied by genome-wide expression profiling using Affymetrix microarrays , namely five control Tcf1+/− mice , four Tcf1−/− mice without tumor , and eight thymic tumors from Tcf1−/− mice . Expressions of several oncogenes , known to be involved in leukomogenesis , were analyzed and were not up-regulated in the thymic lymphomas compared to the Tcf1−/− without lymphomas ( Tal1 , Tal2 , Lyl1 , Lmo1 , Lmo2 , SilTal , p53; unpublished data ) . Analysis of components of the Wnt pathway confirmed that Tcf1 ( Tcf7 ) expression was absent ( as expected ) in the Tcf1−/− thymocytes ( with and without a tumor ) , whilst in all but one Tcf1−/− tumor sample , the expression level of the transcription factor Lef1 was up-regulated compared to control ( Tcf1+/− ) thymocytes ( Figure 3A , left panel ) . Principal component analysis of the Wnt target genes in all 17 thymic samples confirmed the obvious discriminating factor between Tcf1−/− and Tcf1+/− samples to be Tcf7 ( the HUGO gene name for Tcf1 ) . The Tcf1−/− tumor samples were clearly distinguished by factors involved in the Wnt-signaling pathway , Axin2 , Lef1 , and Tnfrsf19 , or in the Notch signaling pathway , Deltex1 and Hes1 ( Figure 3A , right panel ) . These results indicated that both Wnt and Notch signaling are affected in the Tcf1−/− tumor samples compared to the other two groups . Tcf1−/− samples without a tumor were distinguished by low expression of the following factors: Emp1 , Krt8 , Runx2 , CD44 , Fn1 , Jag1 , Id2 , and Cdh1 . Several of these genes are known to be Wnt target genes ( Runx2 , Id2 , CD44 , and Fn1 ) . These data show ectopic up-regulation of Wnt signaling as demonstrated by high expression of Lef1 , CyclinD1 , and c-Myc as well as Notch target genes Hes1 and Deltex1 ( Figure 3B ) . Collectively , these data indicate that interaction between the Wnt and Notch pathways is necessary for full lymphomagenesis . Confirmation of the array data was performed with a panel of 40 Tcf1−/− thymic lymphomas by Q-PCR . In all tested tumor samples , the expression level of Lef1 was increased compared to thymocytes of control mice ( Figure 3C ) . The mean expression Axin2 level of the 40 tumor samples was 4 times elevated compared to the mean expression Axin2 level of the control mice ( 1 . 2 versus 0 . 3 ) , with 29 of the 40 tumor samples ( 73% ) having a higher Axin2 level than 0 . 3 ( Figure 3C ) . Moreover , the high Axin2 expression in the majority ( 73% ) of lymphomas in combination with the universally up-regulated Lef1 expression indicates a marked increase in Wnt signaling in these lymphomas . Further analysis of this panel of lymphomas showed that the expression levels of Hes1 and Deltex1 , two target genes of Notch1 signaling , were enhanced in all tumor samples compared to the control samples ( Figure 3C ) , again demonstrating that both the Wnt and Notch pathway are involved in full lymphomagenesis . As high Lef1 expression is already present in pre-leukemic samples ( Figure 3A ) , it is likely that deregulated Wnt signaling predisposes thymocytes to induction of activating somatic mutations in Notch1 , which subsequently accelerate lymphoma development . To confirm the paradoxical finding that mice lacking Tcf1 suffer from thymic lymphomas due to deregulated high Wnt signaling rather than low , we crossed Tcf1−/− mice with a well-established Wnt-reporter mouse strain , namely the Axin2-LacZ mice . Wnt-activity in these mice can be measured by the expression of β-galactosidase driven by the Axin2 promoter . In Figure 4A , the CD4/CD8 dot plots are shown of thymocytes of four different representative mice . The histograms show the Wnt-activity in DP , ISP , and DN3 cells for Tcf1+/− thymocytes ( filled ) , Tcf1−/− thymocytes ( thin line ) , and tumor Tcf1−/− thymocytes ( thick line ) . The thymocyte subsets of a Tcf1−/− control mouse without a tumor show severely reduced Wnt-activity in ISP and DN3 thymocyte subsets compared to the Tcf1+/− control mouse ( Mean Fluorescence Intensity [MFI] of 385 and 104 compared to 874 and 635 in control ISP and DN3 , respectively ) , indicating a strongly diminished nuclear response to Wnt signals due to the Tcf1 deficiency . Interestingly , residual Wnt-activity can be measured in Tcf1−/− thymocytes , which suggests that Lef1 is mediating low levels of Wnt-activity in Tcf1−/− mice as a likely compensatory mechanism ( as also shown by Figure 3A ) . Tcf1−/− mice developing lymphomas show enhanced Wnt activity in the developmental stages in which the tumor cells are blocked ( MFI of 1 , 425 and 1 , 225 for Wnt-reporter signal in DP and ISP for tumor 1 and 2 , 123 , 2 , 374 , and 1 , 203 in DP , ISP , and DN3 for tumor 2 ) . The thymi of the Tcf1+/− control mouse and the two Tcf1−/− tumor mice displaying high Wnt activity were further examined for the RNA expression levels of Lef1 and Hes1 . The expression level of Lef1 was increased in both Tcf1−/− induced lymphomas , indicating that these high levels of Lef1 underlie the highly active Wnt signals in these tumors ( Figure 4B ) . Interestingly , only in tumor 2 ( >175×106 cells ) were high levels of the Notch target gene Hes1 observed , indicating that Notch signaling accelerates or maintains tumor development once it is initiated by deregulated Wnt-signaling . Indeed , when we compared the thymus size to the expression level of Hes1 , we found that only in large Tcf1−/− tumors ( >25×106 cells ) is the expression level of Hes1 increased ( Figure 4C ) . These data suggest that a first oncogenic hit is the deregulation in Wnt-signaling due to high levels of Lef1 and that deregulation of the Notch-pathway is a secondary acquired mutation . To check for mutations in Notch1 , we sequenced both the heterodimerization domain ( HD ) , exon 26 and 27 , and the PEST domain , encoded by exon 34 . This analysis showed that in Tcf1−/− thymi ( without tumor ) and Tcf1+/− samples , no mutations were found in the three exons ( n = 8 , unpublished data ) . Analysis of the panel of 40 Tcf1−/− thymic lymphomas demonstrated mutations in exon 34 in all but one thymic lymphoma sample ( unpublished data ) , which is known to promote Notch1 signaling by increasing the half-life of intracellular Notch , hence promoting tumor survival and growth . Together , these results suggest that Tcf1 deficiency leads to a pre-leukemic stage that favors additional mutations , most notably in Notch1 . To gain further insight into the mechanism underlying lymphomagenesis , the balance between the long and short isoforms of Lef1 and Tcf1 was investigated . It is known that the balance between the long and short isoforms of these factors is crucial in regulating Wnt signaling , as only the long isoform can bind β-catenin and hence mediate Wnt signaling , whilst the short isoform is considered the natural antagonist of Wnt signaling ( a simplified version of the short versus the long form of Lef is shown in Figure 5A ) . Analysis of the Tcf1−/− tumor samples ( RNA and protein ) revealed that the normal ratio of long over short isoforms for Lef1 was altered in favor of the long isoform of Lef1 , which mediates transcription of Wnt-β-catenin target genes ( Figure 5A ) . The RNA levels of Tcf1 and Lef1 in normal thymus indicated a 10 times higher expression of Tcf1 than Lef1 in the thymus ( Figure 5B ) . This is further illustrated by an example of the protein levels of Tcf1 and Lef1 in a nuclear extract protein sample of total thymocyes ( Figure 5B , right panel ) . Reciprocal regulation of Tcf1 and Lef1 at the protein level was further examined in the major sorted thymic subsets ( DN , DP , and SP ) . Of interest , the ratio between long Wnt responsive isoforms and short repressor isoforms is different for Tcf1 versus Lef1 . For Tcf1 , there is a clear expression of the inhibitory short isoforms in all thymocyte subsets , whilst for Lef1 in all stages the long β-catenin binding form is more abundant , except for the single-positive stage ( Figure 5C ) . This suggests that major repressors of Wnt signaling in the thymus are formed by the Tcf1 short isoforms . Therefore , in the absence of Tcf1 , a repression of Lef1 expression in the thymus is diminished . This allows for high levels of Lef1 , which naturally occurs in a ratio of more Wnt responsive than inhibitory isoforms , hence strengthening the Wnt responsiveness of the ( pre ) leukemic cells . Thus , a major function of Tcf1 appears to control Lef1 expression via its short isoforms . The data described above suggest that both deregulated Wnt as well as Notch signaling are required for development of the Tcf1-deficient lymphomas . To investigate the Wnt and Notch dependency , we performed a number of experiments with pharmacological drugs and genetic tools . Using the γ-secretase inhibitor DAPT , a potent Notch1 inhibitor , clear loss of survival and proliferation was observed in Tcf1−/− tumor cells ( Figure 6A , left figure ) . Moreover , the high Lef1 levels in the tumor cells appeared not to be a result of the deregulated Notch signaling , as inhibition of the Tcf1−/− tumor cells by DAPT only mildly affects Lef1 levels , whilst both Notch target genes Hes1 and Deltex were completely down-regulated ( Figure 6A , right figure ) . Ongoing activation of the Wnt signaling pathway was shown to be crucial for the survival of the Tcf1−/− tumor cells in two different sets of experiments . First of all , incubation of the generated Tcf1−/− cell lines with the Wnt inhibitor Quercetin , which blocks the interaction between β-catenin and Tcf/Lef factors , induces rapid cell death ( 7 h ) , whilst the non-Wnt-dependent Jurkat cell line was minimally affected ( Figure 6B ) . In addition , transfection of a dominant negative form of Tcf1/Lef1 in Tcf1−/− lymphoma cells also induced rapid cell death ( >80% cell death after 6 h; Figure 6C ) . Thus , both ongoing Wnt and Notch signaling are required for the survival of Tcf1-deficient lymphomas . To further investigate the Wnt and Lef1 dependency , we performed reporter gene analysis using the natural Lef1 promoter , which contains four consensus Tcf/Lef binding sites ( Figure 6D ) . The luciferase experiments with the Lef1 promoter show that this promoter is Wnt responsive and demonstrate that Lef1 expression can be up-regulated by β-catenin-Lef1 complexes , providing a positive feedback loop . Such a positive feedback loop has been suggested before [25] . Consistent with the Wnt responsiveness , transfection of a ΔN Tcf1 construct , which acts as a dominant negative competitor , was capable of abolishing the β-catenin-induced activation of Lef1-dependent transcription . Thus , the Lef1 promoter is Wnt responsive and negatively regulated by short Tcf1 isoforms that lack the β-catenin interaction domain . Taken together , these data indicate that the highly deregulated Wnt signaling in the tumor cells is driven by Lef1; that it is frequently associated with increased Notch signaling , which acts as a collaborative oncogenic event; and that it is continuously required for survival of these lymphomas .
During normal hematopoiesis , Tcf1 is required to induce and maintain proliferation of developing T cells in the thymus [19] and inhibit apoptotic signals at the DP stage [23] . Absence of Tcf1 not only results in a severely reduced thymic cellularity but also blocks differentiation of the thymocytes [11] , [19] , [21] . It is remarkable that Tcf1−/− cells blocked in differentiation develop into lymphoma cells , as we here report . We propose that Tcf1−/− blocked thymocytes give rise to lymphoma cells due to deregulated Wnt signaling , which is driven by expression of deregulated expression of Lef1 . This model is based on several key observations ( Figure 7 ) . First of all , in essentially all Tcf1−/− tumors , a high expression of Lef1 ( selectively of the long form of Lef1 , which is able to bind β-catenin ) was found . This up-regulation of Lef1 likely acts as a compensatory mechanism for the lack of Tcf1 and is probably caused by lack of repression by Tcf1 . We show much higher Tcf1 than Lef1 expression in the normal thymus , including the naturally occurring dominant negative isoforms of Tcf1 . This suggests a direct repressor function of Tcf1 for Lef1 expression given the normally much higher Tcf1 than Lef1 expression in thymus . Especially since Lef1 has more long-beta catenin responsive isoforms ( except in the SP stage ) , while Tcf1 has slightly more expression of the short form . Hence in the complete absence of Tcf1 , Lef1 will take over as a Wnt-responsive transcription factor in the thymus . Thus , the lymphoma development is initiated by developmental arrest due to lack of Tcf1; thereby , the suppression of Lef1 expression by short Tcf1 isoforms is lifted . This leads to higher Lef1 expression and a propensity to higher Wnt responsiveness , restoring proliferation and increasing Wnt target gene expression and concomitantly to the possibility of induction of somatic mutations , such as those found in Notch1 . Lef1 expression can be further enhanced by Notch , as shown in lymphomas that lack the E2A transcription factor [27] . Moreover , Lef1 can also positively regulate its own expression through Wnt dependent ( [26] , plus data in Figure 6D ) and independent mechanisms [26] , [28] . Finally , up-regulation of other oncogenes will lead to frank lymphoma/leukemia development . In this respect , up-regulation of T-ALL oncogenes such as Lmo2 and Mef2C in the Tcf1−/− thymocytes may also contribute to the preleukemic nature of these cells . Importantly , the in vivo evidence for this model was provided by crossing Tcf mice with Axin2-LacZ-reporter mice . Tcf1−/− mice without tumors have a reduced level of Wnt-activity in all thymocyte subsets compared to Tcf1+/− mice , whilst Tcf1−/− thymic lymphoma cells show a very high level of Wnt-activity in blocked thymocytes . It is of interest to compare the Tcf1−/− lymphomas with two other murine lymphoma models , namely those induced by activated β-catenin and by lack of E2A . The development of Tcf1−/− lymphomas contrasts with thymic lymphomas induced by overexpression of β-catenin in that the latter show no Notch1 mutations [29] but may be dependent on p53 absence [30] . In the E2A-deficient lymphomas , Lef1 is a Notch target gene only in the context of the lymphoma cells , but not in normal progenitors , and Lef1 is essential for lymphomagenesis . Thus , in the absence of normal regulatory mechanisms in thymic lymphoma cells provided by E2A or Tcf1 , Lef1 can act as an important oncogene . Interestingly , the Tcf1−/− thymic lymphomas easily gain additional mutations in the Notch1 gene ( in contrast to the β-catenin-dependent lymphomas ) , which leads to further development of these lymphomas . Once Notch1 expression is established , it may serve as an accelerator of the Lef1-mediated deregulated Wnt signaling , ensuring increased survival and expansion of the lymphoma cells [27] . In the human equivalent of these T-cell lymphomas , T-cell acute lymphoblastic leukemias ( T-ALL ) , several genetic abnormalities have been described including Notch1 mutations in a large proportion of all human T-ALL [31] . While it is difficult to unravel the stepwise process of leukemia development in humans , the activating mutations in Notch1 are not always the initiating events as shown by data from a leukemia that was observed in a gene therapy trail for X-linked SCID [32] . In this case , it was conclusively shown that insertional mutagenesis near the LMO2 proto-oncogene was the first genetic aberration followed by Notch1 mutations and further genetic aberrations [32] . It will be of high interest to see if loss of Tcf1 tumor suppressor function occurs in human T-ALL . Whether loss of function of Tcf1 as a tumor suppressor gene actually occurs in human T-ALL is currently under investigation . Human T-ALL with mutations in Lef1 have been described , although the mechanistic consequence of these mutations is currently unclear [33] . Two recent studies from the Bhandoola and Gounari laboratories , respectively , have conclusively demonstrated a key role for Tcf1 in establishing T-cell commitment [10] , [16] . Deletion of β-catenin in the thymus has been reported both to affect T-cell development ( using Lck-CRE ) or to have no effect at all , when uisng Mx-Cre-mediated deletion of β-catenin [34] or β- and γ-catenin simultaneously [35] , [36] . However , our recent work on Wnt dosage in various hematopoietic lineages including thymocytes suggests that the lack of phenotype using mx-Cre might be caused by the fact that Wnt signaling was not completely abolished in these models [37] , [38] . Moreover , the Held Group also published that the Tcf1 phenotype can be complemented by transgenic expression of a long Tcf1 isoform but not by a short ( non/Wnt responsive ) isoform [23] . Therefore , a major role for Tcf1 in the thymus is to integrate Wnt responsive signals and thereby allow T-cell development to occur normally . Nevertheless , our current work also indicates an important non/Wnt-dependent role of TCF , as a negative regulator of the Wnt pathway . The recent work of Bhandoola is also interpreted as a non/Wnt-dependent effect . Therefore , it is an intriguing possibility that both Wnt-dependent effects of Tcf1 ( e . g . , Wnt-driven proliferation of DN thymocytes ) and Wnt-independent effects ( induction of T-cell specification ) collaborate in the early stages of T-cell development . In summary , we here report that Tcf1 has a dual function during T-cell development: first , it is needed as a transcriptional activator of Wnt-induced proliferation , but unexpectedly it also acts as a transcriptional repressor and tumor suppressor gene to prevent the development of thymic lymphomas and it may also function in a Wnt-independent way in these early stages of T-cell development , as suggested by recent work [16] perhaps by repressing genes of alternative ( non-T ) lineages . We conclude that Tcf-1 acts as a molecular switch between proliferative and repressive signals during T-cell development in the thymus .
C57Bl/6 Tcf1−/− ΔVII/ΔVII were originally described by Verbeek [19] , C57Bl/6-CD45 . 1 ( Ly5 . 1 ) and C57Bl/6-Rag1−/− mice were obtained from the Jackson Laboratory , and Conductin ( Axin2 ) -LacZ mice were kindly provided by B . Jerchow and W . Birchmeier ( Max Delbrück Center for Molecular Medicine , Berlin , Germany ) [39] . All mice were kept in the specified pathogen-free ( SPF ) breeding section , and this study was approved by the institutional Animal Ethical Committee of the Erasmus MC , Rotterdam and the Leiden University Medical Center , Leiden . Paraffin sections of organs of Tcf1−/− mice were stained with H&E or with antibody against CD3 ( A045229; DAKO , Glostrup , Denmark ) and biotinylated goat anti-rabbit IgG ( BA-1000; Vector Labs , Burlingame , CA , USA ) as the secondary antibody . Visualization was enforced with ABC staining kit ( PK6100 , Vector Labs ) with 3 , 3′-diaminobenzidine tetrahydrochloride ( DAB , D5637 , Sigma-Aldrich , St Louis , MO , USA ) as substrate . Mayer's hematoxilin was utilized as nuclear counterstaining . The following antibodies were obtained from BD Biosciences ( San Diego , CA ) : anti-CD3-APC ( 145-2C11 ) , anti-CD4-PeCy7 ( RM4-5 ) , anti-CD8-PerCP ( 53-6 . 7 ) , anti-CD25-PE ( PC61 ) , anti-CD44-PE ( IM7 ) , anti-CD24-FITC ( M1/69 ) , anti-cKit-PeCy7 ( 2B8 ) . Lineage markers Mac1 ( M1/70 ) , Gr1 ( ( RB6-8C5 ) , B220 ( Ra3-6B2 ) , and Ter119 and Nk1 . 1 ( PK136 ) were all biotinylated and streptavidin APC-Cy7 . Cells were stained in Fluorescence-activated cell sorter ( FACS ) buffer ( PBS , 2% bovine serum albumin , 0 . 1% sodium azide ) for 30 min at 4°C . Intracellular β-galactosidase activity was measured by staining cells with 2 mM fluorescein di-β-D-galactopyranoside ( FDG ) substrate ( Molecular Probes ) . FDG was loaded into the cells by hypotonic shock at 37°C for 1 min , prior to cell surface antibody staining . The β-galactosidase reaction was stopped with 1 mM phenylethyl β-D-thiogalactopyranoside ( PETG , from Molecular Probes ) . Cells were washed and immediately analyzed on a Canto I ( BD Biosciences ) . Data were analyzed using FlowJo software ( Tree Star , Ashland , OR , USA ) . Lymphoma or thymocyte suspensions of previously characterized mice were prepared aseptically . Cells ( 5×105 ) were injected in the tail vein of sublethally irradiated ( 4 Gy ) Rag1−/− mice . The mice were bled every 4 wk until the end of the experiment . The presence of lymphoma cells was investigated by flow cytometry . DNA ( 10 µg ) was digested with EcoRI overnight at 37°C , separated on a 0 . 7% agarose gel , and blotted onto a positively charged nylon membrane ( Hybond , Amersham ) . Southern blots were probed with a 32P-labeled 1 . 2 kb EcoRI-ClaI genomic fragment recognizing the Jβ2 region of the TCRβ gene . Total protein lysates and nuclear extracts were generated from total thymocytes and sorted populations ( DN , DP , and SP ) . Total protein lysates were generated by immediately lysing the cells in boiling sample buffer ( 150 mM Tris-HCl pH 6 . 8 , 300 mM DTT , 30% glycerol , 6% SDS , 0 . 1% bromophenol blue ) . Nuclear extracts were generated by resuspending cells in buffer A ( 10 mM Hepes , 1 . 5 mM MgCl2 , 10 mM KCl , 1 mM EDTA , 0 . 5 mM DTT and freshly added protein inhibitor cocktail [PIC] ) for 15 min on ice . Subsequently NP40 ( final concentration 0 . 6% ) was added , thoroughly mixed , and the cytoplasmic extract was removed by centrifugation . Remaining nuclei were lysed by incubating with buffer C ( 20 mM Hepes , 10% glycerol , 1 . 5 mM MgCl2 , 0 . 42 M NaCl , 0 . 2 mM EDTA , and freshly added DTT and PIC ) for 30 min at 4°C . Nuclear extracts were ready after centrifugation . Protein concentration was measured using BCA Protein Assay kit ( Pierce , Rockford , MD , USA ) . Lysates containing 1 µg of protein were separated by electrophoresis on a 10% SDS-polyacrylamide gel and transferred onto PVDF membranes . Nonspecific binding was blocked by incubation in blocking buffer ( 2 . 5% BSA in TBS-Tween ) followed by incubation with the primary antibodies and the appropriate secondary antibodies conjugated to horseradish peroxidase . All isoforms of Tcf1 were detected by anti-Tcf1 antibody ( clone C46C7 , rabbit mAb , Cell Signaling , Boston , USA ) , and all Lef1 isoforms were detected by anti-Lef1 antibody ( clone C18A7 , rabbit mAb , Cell Signaling ) . Equal loading was confirmed by reprobing the blots with an anti-actin antibody . Thymocytes were homogenized for RNA isolation using Qiagen RNeasy minicolumns . The quantity and quality of total RNA was determined using spectrophotometry ( Nanodrop ) and an Agilent Bioanalyzer . One µg of RNA was used to generate cRNA using Affymetrix One cycle cDNA synthesis kit ( Affymetrix , Santa Clara , CA , USA ) , after which the samples were biotinylated using an Affymetrix IVT labeling kit ( Affymetrix ) . The samples were hybridized overnight at 42°C to GeneChip mouse genome 430 2 . 0 Arrays ( Affymetrix ) . Washing and staining steps were performed on a Fluidics station 450 , and the Genechips were scanned using a GeneChip scanner 3000 ( Affymetrix ) at the Department of Immunology , Erasmus Medical Center . Raw data were normalized and summarized using Robust Multichip Average ( RMA ) method [40] . Array analysis was performed using R-2 . 14 ( http://cran . r-project . org/ ) and Bioconductor 2 . 9 software ( http://www . bioconductor . org/ ) using the bpca [41] and gplots [42] packages . From the dataset , genes were selected for display in heatmaps , in which the rows of the expression matrix were ordered by hierarchical clustering of Eucledian distances between the samples , with the expression intensities being scaled per probeset . Principal component analysis was performed on a dataset of Tcf1+/− , Tcf1−/− , and Tcf−/− tumor samples , using selection of Wnt and Notch response genes ( Emp1 , Tcf7 , Tnfrsf19 , Hes1 , Dtx1 , Notch1 , Axin2 , Lef1 , Cd44 , Runx2 , Fn1 , Cdh1 , Jun , Ccnd1 , Krt8 , Id2 , and Jag1 ) . The first three principal components are displayed . Total RNA was extracted using Qiagen RNeasy minicolumns . One µg of total RNA was used as a template for cDNA synthesis , using Superscript II reverse transcriptase ( Invitrogen , Carlsbad , CA , USA ) , Oligo dT , and random hexamer primers . The RQ-PCR reaction was performed using TaqMan Universal mastermix ( Applied biosystems , Foster City , CA , USA ) and was run on a PRISM 7700 sequence detection system containing a 96-well thermal cycler ( Applied Biosystems ) . The following primers were used in combination with FAM-labeled probes from the universal probe library ( Roche ) : Deltex1 forward primer: 5′-GAAGAACTTGAATGGCACTGG-3′; reverse primer: 5′-GTTTGGGTGCTCGTGTCAG-3′; Lef1 short forward primer: 5′-GCGACACTTCCATGTCCAG-3′; reverse primer: 5′-TCCTGTTTGACCTGAGGTGTTA-3′; Lef1 long forward primer: 5′-TGGTTAACGAGTCCGAAATCA-3′; reverser primer: 5′-AGAGGACGGGGCTTGTCT-3′; Axin2 forward primer: 5′-GCAGGAGCCTCACCCTTC-3′; reverse primer: 5′-TGCCAGTTTCTTTGGCTCTT-3′; Hes1 forward primer: 5′-AAACACTGATTTTGGAGCACT-3′; and reverse primer: 5′-TGCTTCACAGTCATTTCCAGA-3′ . RQ-PCR results were normalized to Abl expression in the same sample: forward primer: 5′-TGGAGATAACACTCTAAGCATAACTAAAGGT-3′; reverse primer: 5′-GATGTAGTTGCTTGGGACCCA-3′; and probe: 5′-FAM-CCATTTTTGGTTTGGGCTTCACACCATT-TAMRA-3′ . cDNA of total thymus was used for the amplification of exons encoding the Notch1 heterodimerization and PEST domains . Primers used for the identification of activating Notch1 mutations are described elsewhere [43] . Several Tcf1−/− cell lines were established from Tcf1−/− thymic lymphomas , and all cell lines show Notch1 mutations and a high ratio of Lef1 long over short isoform . Transfection experiments were performed by transfecting the cell line with eGFP together with either a control construct ( pcDNA3 ) or a dominant negative Lef1/Tcf1 construct ( transfection ratio GFP∶construct , 1∶10 ) using AMAXA electroporation technology . Transfected cells were identified based on GFP positivity and phenotype , and cell viability was determined 6 h after transfection . Discrimination between viable and dead cells was performed by staining the cells with AnnexinV and 7AAD ( BD Bioscience ) . Tcf1−/− cell line cultures were performed in the presence and absence of the γ-secretase inhibitor DAPT ( 0 , 5 , and 50 µM ) or Quercetin ( 50 µM ) . At the indicated time points , cell cycle analysis was performed using propidium iodide , live cells were determined by 7AAD/AnnexinV stain , and RNA was isolated for gene expression levels . 293 T cells were cultured in Iscove's Modified Dulbecco's Medium supplemented with 10% fetal bovine serum , L-glutamin , and penicillin/streptomycin and transfected using the Fugene method according to the manufacturer's procedures ( Roche ) . The cultures were transfected with 1 . 5 µg LEF-1 4000 luciferase reporter plasmid ( containing four Tcf/Lef-responsive elements ) or 1 . 5 µg of the LEF-1 600 luciferase reporter plasmid ( all Tcf/Lef-responsive elements deleted ) ( kindly provided by Dr . J . Skokowa , Hannover Medical School [44] ) . The cells were cotransfected with S33-βcatenin and/or pCI and/or ΔN-Tcf ( 3 µg ) . To control for transfection efficiency , all transfections included the pRLTK-renilla reporter ( 0 . 15 µg ) . Transfected cells were cultured for 24 h and then lysed and assayed for reporter activity . Luciferase and Renilla activity was measured using a dual-luciferase reporter assay system from Promega ( Madison , USA ) . All luciferase activities were normalized to Renilla activities . Statistical analysis was performed using the Mann–Whitney U test ( Prism GraphPad Software , San Diego , CA , USA ) . p<0 . 05 was considered statistically significant . | Cancers often develop as a consequence of deregulated expression of key factors that operate during normal development . T-cell factor 1 ( Tcf1 ) has an established role in the nuclear response to Wnt signaling during normal T-cell development in the thymus . Here we show in mice that the absence of Tcf1 can trigger tumorigenesis . As expected from previous work , lack of Tcf1 results in a small thymus with several partial blocks in T-cell development in the thymus . Surprisingly , we observe that a large proportion of Tcf1−/− mice spontaneously develop thymic lymphomas . Thorough investigation of these thymic-derived tumors revealed that the mechanism underlying these lymphomas is , paradoxically , increased levels of Wnt-signaling . We propose that Wnt-signaling in these tumors is mediated by up-regulated expression of the Tcf1-homologue , Lef1 , and specifically its long isoform . Furthermore , we have evidence to propose that in a normal thymus , short isoforms of Tcf1 that cannot respond to Wnt signals act as repressors of Lef1-mediated Wnt-signaling . Thus , we propose that Tcf1 has a dual function developing T cells in mice: it functions as a T-cell–specific tumor suppressor gene in addition to its established role as a transcriptional activator of Wnt-induced proliferation . Whether loss of function of Tcf-1 as a tumor suppressor gene actually occurs in human T-cell lymphoblastic leukemias is currently under investigation . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"hematologic",
"cancers",
"and",
"related",
"disorders",
"medicine",
"leukemias",
"hematopoiesis",
"immune",
"cells",
"acute",
"lymphoblastic",
"leukemia",
"t",
"cells",
"immunology",
"biology",
"hematology"
] | 2012 | The Nuclear Effector of Wnt-Signaling, Tcf1, Functions as a T-Cell–Specific Tumor Suppressor for Development of Lymphomas |
Herpesviruses establish a lifelong latent infection posing the risk for virus reactivation and disease . In cytomegalovirus infection , expression of the major immediate early ( IE ) genes is a critical checkpoint , driving the lytic replication cycle upon primary infection or reactivation from latency . While it is known that type I interferon ( IFN ) limits lytic CMV replication , its role in latency and reactivation has not been explored . In the model of mouse CMV infection , we show here that IFNβ blocks mouse CMV replication at the level of IE transcription in IFN-responding endothelial cells and fibroblasts . The IFN-mediated inhibition of IE genes was entirely reversible , arguing that the IFN-effect may be consistent with viral latency . Importantly , the response to IFNβ is stochastic , and MCMV IE transcription and replication were repressed only in IFN-responsive cells , while the IFN-unresponsive cells remained permissive for lytic MCMV infection . IFN blocked the viral lytic replication cycle by upregulating the nuclear domain 10 ( ND10 ) components , PML , Sp100 and Daxx , and their knockdown by shRNA rescued viral replication in the presence of IFNβ . Finally , IFNβ prevented MCMV reactivation from endothelial cells derived from latently infected mice , validating our results in a biologically relevant setting . Therefore , our data do not only define for the first time the molecular mechanism of IFN-mediated control of CMV infection , but also indicate that the reversible inhibition of the virus lytic cycle by IFNβ is consistent with the establishment of CMV latency .
Herpesviruses are characterized by their ability to establish a lifelong latent infection in their natural host and reactivate upon immunosuppression . Cytomegaloviruses ( CMV ) are paradigmatic β-herpesviruses , characterized by strict species specificity , but highly prevalent in numerous mammalian species [1] . Human CMV ( HCMV ) prevalence ranges from 30 to 90% [2] . While primary infection and latency are usually asymptomatic in immunocompetent individuals , immune suppression results in virus reactivation , which is associated with substantial morbidity and mortality . In particular , CMV reactivation may result in allograft rejection , pneumonia or gastroenteritis in recipients of solid-organ and bone-marrow transplants [3] . Understanding the molecular mechanisms involved in the establishment and maintenance of latency is fundamental for developing effective countermeasures to CMV disease in high-risk populations . The human and the murine CMV ( MCMV ) share many biological properties . As such , MCMV infection of mice is a widely used in vivo model of CMV immunity and latency . Latency is characterized by the persistence of silenced virus genomes in the absence of infectious viral particles [4] . Both HCMV and MCMV infect a variety of cell types in their respective host [5] , [6] , but latency seems to be confined to distinct cell types , such as cells of the myeloid lineage [5] , [7]–[10] . While HCMV latency in endothelial cells remains controversial [11] , [12] , strong evidence supports the notion that liver sinusoidal endothelial cells ( LSECs ) are a site of MCMV latency [13] . Similar to HCMV , chromatinization and recruitment of cellular repressors to the viral DNA and to the major immediate early ( MIE ) gene locus are critically involved in the in vivo establishment of MCMV latency [14] , [15] . The IE genes regulated by the MIE promoter ( MIEP ) encode the first viral proteins expressed during productive infection , and act as essential transactivators of early and late genes [16] . Reactivation of latent HCMV from in vivo infected myeloid progenitor cells was shown to be related to MIE chromatinization [17] . Namely , the latent viral genome is associated with repressive chromatin in immature myeloid cells , whereas virus reactivation is accompanied by chromatin remodeling and initiation of transcription at the MIE locus during cell-differentiation . Therefore , MIEP transcriptional activity is generally considered an important checkpoint in CMV latency and reactivation . In the immunocompetent host , primary infection is controlled by a combination of immunological effectors . Infected cells are directly eliminated , e . g . by cytotoxic effects of NK or T-cells . In addition , the spread of infectious virus is restricted by antibodies or by cytokines that reduce the permissiveness of cells for viral replication . Cytokines such as type I ( IFNα/β ) or type II interferons ( IFNγ ) are critical in the control of acute infection [18] , [19] . They exert their antiviral action by activating immune effector cells like DCs , T cells or NK cells , but also by inducing transcriptional programs which suppress virus replication in target cells [20] . While it is generally accepted that interferons limit virus spread without killing the infected cell , the exact mechanism of their antiviral action remains unclear . Most importantly , it remains unclear if their effect results in CMV clearance , or if the viral replication is merely suppressed while genomes are maintained in the infected cell . A reversible block of viral replication prior to immediate-early expression would argue that interferons play a key role in the establishment of CMV latency . In a seminal paper , Presti et al . showed that mice that lack type II IFN receptors maintain a productive MCMV infection and that MCMV reactivation from explants of latently infected mice may not be observed in the presence of IFNγ [19] . Unfortunately , this experimental setting could not differentiate if the IFNγ truly suppresses virus reactivation by acting directly in latent cells or merely inhibited viral spread to other cells upon reactivation . In contrast , the role of type I IFN in the establishment and maintenance of latency is difficult to investigate in vivo , as IFNα/β receptor knockout ( IFNAR− ) mice are about 1000-folds more susceptible to MCMV than wild-type mice , and die within a few days post infection [19] . Nevertheless , in vitro experiments showed that IFNβ induced by lymphotoxin α reversibly suppresses HCMV and MCMV gene expression and replication [21] . Moreover , MCMV replication in macrophages is transiently suppressed by synergic action of IFNγ and type I interferons [22] . However , both publications showed that the suppression was only partial , because viral gene expression was reduced , but still detectable [21] , [22] . Therefore , the effects of this axis seemed to reflect simmering lytic replication , rather than bona fide viral latency . In this study , we show that MCMV replication may be completely , but reversibly , inhibited in cells that respond to IFNβ , in a manner consistent with viral latency . On the other hand , cells which failed to respond to IFNβ were permissive for MCMV replication . We show that the inhibition of MCMV replication by IFNβ depends on the inhibition of viral gene expression at the level of IE transcription mediated by nuclear domain 10 ( ND10 ) components , which is fully reversible even after extended culture of in vitro infected cells and in cultures of endothelial cells derived from latently infected mice . In summary , our data indicate that reversible silencing of viral genomes by IFN-induced ND10 components is a key contributor to the establishment of CMV latency .
LSECs are a site of MCMV latency [13] . We recently described an LSEC line which enters cell cycle in a doxycycline-dependent manner and is highly permissive for MCMV infection [23] . To study type I interferon ( IFN ) effects on MCMV replication in quiescent LSECs , growth-arrested cells were incubated with IFNβ for 24 h , infected with MCMV at a multiplicity of infection ( MOI ) of 0 . 001 and viral growth was assayed for a week . Until 5 days post infection ( dpi ) , infectious virus was only exceptionally detected in supernatants ( SN ) of IFNβ-treated LSECs , and viral titers were substantially diminished on 6 and 7 dpi , as compared to untreated cells ( Figure 1A ) . Therefore , consistent with previous reports , IFNβ treatment resulted in delayed viral growth and reduced viral titers , but did not completely block MCMV replication . Recently , we reported that a minor proportion of cells remain unresponsive even to high doses of type I interferon [24] . We speculated that MCMV infection of IFNβ-unresponsive cells may generate sufficient amounts of virus to overcome the barrier to infection installed by IFNβ pre-treatment . In that case , infection at low MOI would increase the chance that exclusively IFN-responsive cells are infected and that the infection becomes fully contained . We thus repeated the initial experiment with reduced doses of MCMV , up to a dilution of 1 plaque forming unit ( PFU ) per 10 , 000 cells , and monitored the long-term IFNβ effects for up to 19 dpi in growth-arrested LSECs ( 20 , 000 cells/well ) . While IFNβ treated samples infected at an MOI of 0 . 1 and 0 . 01 showed substantial virus titers by 7 dpi and later ( Fig . 1B , upper panels ) , MCMV replication was completely abrogated , when the infection was performed at MOIs below 0 . 01 ( Fig . 1B , lower panels ) . It is important to note that both an MOI of 0 . 001 and 0 . 0001 still resulted in complete cell lysis and high viral titers in IFNβ-naïve samples ( Fig . 1B , lower panels , white bars ) . These findings were consistent with the model that CMV infection is contained at very low MOIs because it is restricted to IFNβ-responsive cells . To confirm this hypothesis , we used reporter cells that express an IRF-7-mCherry fusion protein under the control of the IFNβ-responsive IRF-7 promoter ( Figure 2A ) . Reporter cell stimulation with IFNβ ( 500 U/ml ) revealed a small but notable population of non-responding cells ( Figure 2B ) . We separated the cells into responders and non-responders by fluorescence activated cell-sorting ( FACS ) and then infected them with a dose ( MOI 0 . 01 ) , which could not be contained by IFNβ in the previous experiment ( Figure 1B ) . In the absence of further IFNβ treatment , MCMV titers were diminished in cells which responded to IFNβ ( Figure 2C , left diagram ) . More importantly , virus replication was completely abolished upon continuous IFNβ treatment , but only in IFNβ-responsive cells ( Figure 2C , right diagram ) . In summary , these experiments demonstrate that IFNβ pretreatment is sufficient to restrict MCMV replication in cells which respond to IFNβ . However , virus expansion in the few IFN-unresponsive cells eventually overcomes the resistance of the IFNβ-responsive population . IFNβ abrogated productive MCMV replication in LSECs at low MOIs . To elucidate at which step of infection this block was exerted , we infected LSECs with a recombinant MCMV that expresses two different fluorescent proteins as reporters that reflect the activity of the MCMV major immediate early promoter ( MIEP ) . This virus was named MCMVr and contains an ectopically inserted , full-length MIEP sequence flanked by the yellow fluorescent protein EYFP , driven by the ie1/3 promoter , and the red fluorescent protein tdTomato that is controlled by the ie2 promoter . MCMVr grows like WT MCMV in vitro and expresses EYFP and tdTomato with the same scale and kinetics like the MCMV IE1 and IE2 genes , respectively [23] . To determine the onset of viral infection we monitored EYFP expression , which occurs earlier than tdTomato [23] , in line with reports that the ie1 gene is immediately expressed at high levels during primary lytic infection and reactivation [25] , [26] . MCMVr infection of LSECs resulted in strong EYFP-expression , which was hardly detectable in IFNβ-treated cells ( Figure 3A ) . To determine if IFNβ completely abrogated MCMV IE gene expression , LSECs were infected with MCMVr in 96-well plates , scanned for reporter gene expression , and wells were classified as positive when even a single EYFP-fluorescent cell could be observed within a week of infection ( representative result of an experiment in 12 wells per condition is shown in Figure 3B ) . Different IFNβ concentrations were tested with various MOIs , to assess the conditions that allow the complete suppression of viral genes , and the percentage of positive wells was defined ( Figure 3C ) . 500 and 100 U/mL of IFNβ blocked all viral gene expression in more than 80% of wells at 0 . 001 and 0 . 0001 MOI for 7 dpi , whereas 10 U/mL showed similar activity only at the lower MOI ( Figure 3C ) . Importantly , in wells that showed one single positive cell , the progress of the lytic infection was irreversible and the virus would always spread to nearby cells . We next tested if the suppressive effect of IFNβ on MCMV gene expression is permanent or reversible , by removing IFNβ at 7 dpi and monitoring the cells for additional 12 days . Remarkably , IFNβ removal resulted in viral gene expression ( Figure 3D ) and production of infectious virus ( Figure 3E ) about 7–10 days later , while the percentage of EYFP positive wells remained unchanged ( Figure 3D ) and viral titers undetectable ( Figure 3E ) in wells permanently treated with IFNβ . We concluded that IFNβ reversibly suppresses MCMV replication before or at the time of MIEP-driven gene expression . Of note , the same finding was observed following infection with the γ-herpesvirus MHV68 ( Figure S1B ) . In contrast , vesicular stomatitis virus ( VSV ) , which could be efficiently suppressed with IFNβ , was not able to replicate after IFNβ-retraction ( Figure S1A ) . In conclusion , IFNβ silenced the replication of three different viruses . However , this was only reversible for infections with the two herpesviruses . IFNβ reversibly inhibited MCMV replication and expression of genes driven by the ie1/3 promoter . This could be due to a direct block of IE-gene expression , or effects that occurred during the viral entry into the cells . To test if the reversible suppression by IFNβ occurs after the virus has entered the cell and the viral genomes are delivered to the nucleus , we generated a recombinant MCMV in which the ie1/3 locus is flanked by two loxP sites ( IE1/3flox MCMV ) which results in IE1 and IE3 deletion when the genome is recognized by the Cre recombinase in the cell nucleus ( Figure S2A ) . Importantly , IE1/3flox MCMV replicates in Cre-expressing cells , probably due to rapid MIEP-driven gene expression , which precedes the Cre-mediated deletion of target sites ( Figure 4A and Figure S2B ) . We considered that the Cre recombinase would have sufficient time to excise the IE1 and IE3 genes and abrogate reactivation upon IFNβ retraction , if IFNβ blocked viral gene expression after genome delivery to the nucleus . Cre-expressing MEFs were pre-treated with IFNβ , infected with IE1/3flox MCMV or WT MCMV and the wells were scanned for viral plaques . Viral replication of WT and IE1/3flox MCMV was efficiently blocked in cells which constantly received IFNβ over a time period of 4 weeks ( Figure 4A ) . IFNβ removal at 7 dpi resulted in virus replication in several wells infected with WT MCMV , consistent with the data obtained from MCMVr-infected LSECs ( Figure 3D ) . In contrast , IE1/3flox MCMV failed to replicate upon IFNβ removal from Cre-MEFs , indicating that the IE1/3flox MCMV genomes were exposed to Cre-recombinase in the nucleus , and the deletion of the ie1/3 genes abrogated the ability of the virus to replicate upon IFNβ-removal . Since our data indicated that MCMV genomes are delivered to the cell nucleus in the presence of IFNβ , we assumed that it directly inhibits viral gene expression . To test the ability of IFNβ to impair viral gene expression in absence of virion components , we delivered the MCMV genomes into cells by transfection [27] . Since transfection is less efficient in LSECs than in MEFs ( data not shown ) , we transfected MEFs with the MCMVr bacterial artificial chromosome ( BAC ) and treated them with IFNβ immediately upon transfection . MIEP-driven expression was detected by fluorescence microscopy for EYFP expression . Four days post transfection , EYFP was observed in all wells transfected in the absence of IFNβ treatment . In contrast , EYFP could be observed in only few of the IFNβ-treated wells ( Figure 4B ) . Most importantly , removing IFNβ resulted in the restoration of MIEP activity by day 6 ( Figure 4B ) . To understand if the inhibition of gene expression was exclusive to the MIEP promoter , or to any incoming DNA , fibroblasts were transfected with plasmids expressing the EYFP and tdTomato under the control of the MCMV MIEP or with plasmids expressing reporter genes under the control of other promoters ( SV40 and HCMV ) . The expression of all reporter genes was substantially diminished in IFNβ treated cells ( Figure 4C and data not shown ) , indicating that IFNβ suppresses gene expression in a manner that is not specific for the MCMV MIEP but to any incoming DNA . Together , these data provide strong evidence that IFNβ-mediated reversible suppression of viral replication occurs directly at the level of gene transcription of foreign DNA entering the nucleus . To formally show that IFNβ blocks MCMV replication at the level of gene transcription rather than translation , we analyzed the viral transcriptome of IFNβ-treated LSECs immediately upon MCMV infection . CMV particles carry significant amounts of virion-associated RNA [28] , which , upon delivery to infected cells , impede the detection of de novo synthesized immediate early and early viral transcripts . We therefore metabolically labeled newly transcribed RNA with 4-thiouridine ( 4sU ) , isolated the labeled RNA by thiol-specific biotinylation and streptavidin-precipitation [29] , and deep-sequenced the newly transcribed RNA . To observe the effect of IFNβ at the earliest possible time point after the infection , we adapted the infection protocol and incubated the cells with infectious virus for 5 min only , using an infectious dose that was normalized to match an MOI of 10 in standard infection and virus absorption . This allowed us to focus our analysis on viral transcripts generated during the first hour of infection ( hpi ) . At 1 hpi , the IE gene transcripts were detectable and comprised the majority of viral transcripts , whereas they were highly diminished in IFNβ-treated LSECs ( Figure 5A and Table S1 ) . It is important to note that IFNβ treatment also diminished all other viral transcripts that could be detected at 1 hpi , consistent with the observed global suppression of reporter gene expression in all tested expression plasmids . Thus , IFNβ acts at the level of MCMV gene transcription , resulting in strong transcriptional repression of all viral genes expressed in the first hour of infection . Since IFNβ inhibited MCMV replication at the level of viral IE gene transcription we hypothesized that this effect might be mediated by induction of nuclear domain 10 ( ND10 ) components . ND10 bodies are nuclear structures known to associate with incoming viral DNA restricting CMV replication [30]–[32] . Hence , we screened the host-cell transcriptome for members of the ND10 and compared their transcriptional level in untreated and IFNβ-treated LSECs . Interestingly , three major components of the ND10: Daxx , Sp100 and PML , were upregulated in IFNβ-treated LSECs ( Figure 5B and Table S2 ) , consistent with published data [33] , [34] . This was confirmed by immunofluorescence staining for Daxx ( Figure 5C ) and RT-PCR for all three components ( Figure S3 ) . To define the relevance of these factors in the IFNβ-mediated suppression of MCMV replication , we performed shRNA-mediated knockdowns of these three ND10 components , which reduced their mRNA levels to those seen in IFN-untreated cells ( Figure S3 ) . More importantly , each of the three knockdowns was sufficient to almost completely restore MCMV replication in the presence of IFNβ ( Figure 5D ) . Collectively , these data highlight a key role of ND10 bodies in the transcriptional silencing of CMV gene expression induced by IFNβ . While our results showed very clear IFNβ effects on MCMV lytic replication , it remained open if it also induces MCMV latency in vivo . To understand how MCMV infection influences the production IFNβ in vivo , we infected transgenic mice which carry a luciferase reporter gene under the control of the MX2 promoter [35] , a well-characterized IFN stimulated gene ( ISG ) . Luciferase activity could be detected in the MCMV-infected mice already at 4 hpi ( Figure 6A ) , indicating immediate production of IFN upon MCMV infection ( Figure 6A ) . The reporter gene signal peaked at 12 hpi and declined thereafter , although a robust luciferase signal could still be detected by 72 hpi ( Fig . 6A , 6B ) . Interestingly , the response to IFN was most prominent in the liver region throughout the time of monitoring . To determine if in vivo IFNβ-stimulation also transiently silences MCMV in LSECs , we infected mice with MCMVr in which the production of IFNβ was induced prior to infection . For this , we used a previously described IFN-β reporter mouse ( IFN-β+/Δβ-luc ) , allowing the visualization of IFNβ expression by in vivo imaging using firefly luciferase as a reporter [36] . These mice were stimulated with poly I:C and a high activity of the IFNβ promoter could be detected 4 h after poly I:C injection but not in mock treated mice ( Figure S4 ) , consistent with the kinetic of MCMV infection ( Fig . 6A ) . Mice were infected with MCMVr at 8 h post stimulation , and LSECs were isolated from the liver of the infected mice at 72 hpi . LSECs were cultivated for 7 days and analyzed for reporter gene expression at 1 , 4 and 7 days post isolation . After 1 day , MCMV reporter gene expression ( EYFP ) could be detected in all of the wells with LSECs that were isolated from control mice , infected in the absence of poly I:C . In contrast , MCMV ie gene expression was absent in about 1/3 of the wells containing LSECs from poly I:C-stimulated mice ( Figure 6C ) . This was not merely a random redistribution of the EYFP+ cells to fewer wells , because the overall number of EYFP+ cells was substantially reduced in LSECs from poly I:C treated mice ( Figure 6D ) . We considered the possibility that the absence of viral gene expression upon in vivo IFN induction is a result of a hindered viral entry in the LSECs . However , this scenario seemed unlikely , because we could not detect any infectious MCMV in the supernatants from poly I:C pre-treated LSEC ( Figure 6E ) , while control LSECs showed detectable titers , probably as a result of ongoing virus shedding in the first 24 hours of culture . To understand if the in vivo MCMV suppression by IFN was due to reversible silencing of gene expression , the cells isolated from poly I:C-treated mice were further cultivated and MCMV gene expression was monitored at 4 and 7 days post isolation and EYFP expression could be observed in all of the wells , including those that were negative at 1 day post isolation ( Figure 6F ) . Likewise , the number of EYFP+ cells increased upon cultivation , and by 4 and 7 days post isolation the LSECs from poly I:C treated mice showed similar levels as the controls ( Figure 6G ) . Finally , this was accompanied by full virus reactivation , as demonstrated by the emergence of infectious virus in the supernatants at 4 days post isolation in the IFNβ-stimulated LSECs , and by its expansion by day 7 ( Figure 6H ) . In conclusion , the infection of LSECs stimulated with IFNβ in vivo increased the proportion of cells that contained silent MCMV genomes that were able to re-initiate the replication cycle after explantation , upon several days of ex vivo cultivation . To confirm that this also occurs in the course of natural infection , in wild type mice and in absence of poly I:C treatment , we isolated LSECs from BALB/c mice at 72 hpi infection with MCMVr and monitored EYFP expression on day 1 and 4 post isolation . By seeding LSECs at a lower concentration per well ( 50 , 000 cells per well , instead of 70 , 000 ) , and using mice expressing IFNβ from both of its alleles ( luciferase expression in reporter IFN-β+/Δβ-luc mice is possible due to a monoallelic exclusion of IFNβ expression ) , we established conditions where MCMV IE gene expression was completely abrogated in absence of poly I:C prestimulation , because 7 out of 26 wells showed no EYFP expression at all on day 1 post isolation . The majority of these wells ( 5 out of 7 ) became positive for EYFP by day 4 post isolation ( data not shown ) , demonstrating that viral genomes , but no gene expression , were present in some cells immediately upon infection . These silenced genomes may re-initiate the lytic gene expression program , therefore strongly arguing that viral latency is established in parallel with lytic replication at the onset of the in vivo infection . IFNβ reversibly silenced MCMV gene expression in LSECs infected in vivo and in vitro , a phenomenon with intriguing homologies to MCMV latency and reactivation . To determine if IFNβ would be sufficient to suppress MCMV reactivation from LSECs carrying latent viral genomes , and to define if this would also occur at the level of immediate-early gene expression , we infected mice with MCMVr and isolated the LSECs from the liver of latently infected mice . Infectious MCMVr is completely cleared from liver by 14 dpi [23] . Primary LSECs were isolated at 4 weeks post infection and cultivated for up to three weeks in vitro . Viral gene expression was monitored by fluorescence microscopy for EYFP expression . After 6 days of cultivation , the LSEC explant monolayers displayed single fluorescent cells ( Figure 7A ) . Within a couple of days , the infection expanded resulting in numerous fluorescent cells . The majority of wells with LSECs that were cultivated in presence of IFNβ exhibited no viral gene expression ( Figure 7B ) . In contrast , IFNβ removal at 6 days post LSEC isolation resulted in a strong increase of EYFP-positive wells , almost to levels seen in the IFN-naive controls ( Figure 7B ) , thereby excluding the suppressive effects of IFNβ to be due to toxic effects . Finally , infectious virus shedding in the cell supernatants was confirmed only in IFN-untreated cells or upon IFNβ removal ( Figure 7C ) . In summary , these data demonstrate that IFNβ is not only able to efficiently inhibit lytic MCMV infection following pre-treatment , but can also efficiently suppress MCMV reactivation of latently infected primary LSECs .
It is well-established that IFNβ inhibits lytic CMV replication , but cannot abrogate it completely [37] , [38] . Recently , we reported that even high doses of IFNβ cannot activate all cells of a population , leaving a few cells unprotected [24] . We now show that this minority of cells is responsible for the failure of IFNβ to completely abrogate lytic MCMV replication . Consequently , MCMV gene expression and replication are completely blocked by IFNβ at very low doses of infection , when the probability of infection of an IFNβ-unresponsive cell is minimized . Higher doses of infectious MCMV are blocked when infecting sorted IFN-responder cells ( Fig . 2 ) . Restricting the infection to IFN-responsive cells allowed us to identify the reversible nature of the IFN-mediated inhibition of CMV replication . This could not be observed in previous studies , because viral IE gene expression in a single cell is sufficient to overcome IFNβ-induced resistance to viral replication in subsequent rounds of infection . This all-or-nothing phenotype is consistent with a model where the initial failure to contain the expression of IE1 results in a positive feedback loop , which reinforces viral transcription that can no longer be controlled by IFNβ [39] , [40] . Large amounts of virus released from a single IFN-unresponsive cell are then capable of overcoming the antiviral state in the neighboring IFN-responder cells explaining the inability of IFN to fully suppress productive CMV infection in cell culture . Consistent with this model , expression of the viral IE1 protein is crucial for the dispersion of ND10 bodies , thereby allowing transcription of viral early genes to proceed [41] . Several components of the ND10 bodies , are induced by IFN . ND10 were initially described as the nuclear domains where HCMV genomes are localized immediately upon infection [42] . Subsequent studies revealed that HCMV replication is inhibited by additive effects of ND10 components , including PML , Daxx [43] and Sp100 [44] . Daxx has been shown to be involved in chromatin modification [45] , [46] and was found to bind to the MIEP of MCMV in latently infected mice [14] . In addition , a role in transcriptional suppression has been suggested for the nuclear antigen Sp100 which was shown to repress the transcriptional activity of herpes simplex virus 1 ( HSV-1 ) promoters [47] . We showed here that IFNβ-mediated inhibition of MCMV replication critically depends on the ND10 proteins PML , Daxx and Sp100 , rather than on any other IFN induced gene ( Figure 5 ) . Our study supports a critical role of ND10 bodies in limiting the viral transcription at the earliest stages of infection , and shows for the first time that this is fully reversible , and thus consistent with the molecular definition of latency . Therefore , we propose that the virus exploits the IFN-mediated induction of ND10 body components to establish latent infection in tissues strongly responding to IFN . In this context , herpesvirus latency may be understood as an immune evasion mechanism to high levels of IFN , because latency offers a choice for the virus to maintain its ability to reactivate in an environment with rampant immune responses until these responses decline . Our data highlight a crucial role of IFNβ-mediated induction of ND10 components , similar to previous data showing the critical role of PML in the IFN repression of HSV-1 replication [48] . Our results are not necessarily limited to the establishment of latency in endothelial cells . Similar results have been recently observed in macrophages , where interferon induced ND10 expression and an MCMV refractory state at the IE expression level ( M . Hassim and P . Ghazal , personal communication ) . Therefore , IFN may also be involved in the induction of latency in myeloid cells , and it is an intriguing possibility that this may also depend on the induction of ND10 bodies . How do the ND10 bodies silence the viral transcription ? Our results may imply that the silencing is not based on the suppression of a specific promoter , but rather of any incoming episomal DNA , although this still needs to be formally confirmed . More importantly , our results showed that a complex nuclear machinery is required for MCMV silencing , because each of the shRNA knockdowns ( Daxx , Sp100 or PML ) were sufficient to rescue viral transcription , at least in part . Taken together , these results may imply that ND10 bodies silence viral transcription in a manner akin to programmed epigenetic control , but this hypothesis would need to be tested in a detailed study , which goes beyond the scope of this article . IFNβ was not only able to completely inhibit lytic MCMV replication in vitro and in vivo , but also to prevent virus reactivation from latency in explant cultures . Since both IFNα and IFNβ signal through the same receptor and induce a range of similar genes , it is possible that both type I IFNs exert similar effects on MCMV latency [38] , [49] , [50] . On the other hand , recent evidence showed distinct differences in the downstream signaling induced by IFNβ and IFNα [51] . Furthermore IFNβ induces the secretion of IFNα in mice [52] and therefore it is possible that in our experiments the IFNβ-stimulation does not act directly , but rather by enforcing the secretion of other antiviral cytokines which may influence MCMV latency . Either way , it is unlikely that the amounts of type I IFNs which are necessary to keep the virus in check in vitro are produced over a prolonged time in the latently infected host , and this is also inconsistent with our kinetic monitoring of IFN responses upon MCMV infection ( Fig . 6A and 6B ) . However , it is conceivable that individual LSECs which respond to type I IFNs generate a reservoir of latently infected cells . Once viral latency has been established , immune control may well be exerted by primed T and NK cells [53] . These cells are activated later during the infection process , but persist longer than type I IFN secreting cells and both have the potential to secrete IFNγ , and thus control lytical CMV replication [19] . MCMV specific effector T-cells are readily detectable in organs of latently infected mice [54] , arguing for a strong and ongoing recruitment of immune cells to sites of virus latency , and thus for an active role of the immune system in the prevention of CMV reactivation . An additional layer of control may also result from epigenetic silencing of the viral genomes once latency has been established [10] , [15] . In line with a model of epigenetic control of viral transcription , which acts on top of IFNβ mediated transcriptional suppression , IE gene expression restarted with a delay , and could only be observed approximately one week upon IFNβ retraction ( Fig . 3D , 7B ) . In conclusion , our study establishes a link between type I IFN signaling , ND10 bodies and reversible suppression of CMV transcription and strongly argues for their key role in the establishment of herpesviral latency .
All animal experiments were performed in compliance with the German animal protection law ( TierSchG BGBI S . 1105; 25 . 05 . 1998 ) . The mice were handled in accordance with good animal practice as defined by FELASA and GV-SOLAS . All animal experiments were approved by the responsible state office ( Lower Saxony State Office of Consumer Protection and Food Safety ) under permit number 33 . 9-42502-04-11/0426 . M2-10B4 ( CRL-1972; ATCC ) and NIH 3T3 fibroblasts ( CRL-1658; ATCC ) were maintained in Dulbecco's modified Eagle medium supplemented with 10% fetal calf serum , 1% Glutamine and 1% Penicillin/Streptomycin . Primary C57BL/6 , IFNAR−/− and CMV-Cre MEFs were prepared and maintained as described previously [55] . Conditionally immortalized LSECs were generated and cultivated as described [23] . NIH3T3 IRF7-mCherry were generated and characterized previously [24] . All mice were bred at the animal facility of the Helmholtz Centre for Infection Research ( HZI ) and maintained under specific pathogen-free conditions . Conditional deletion/reporter mice IFN-βfloxβ-luc and Mx2Luc reporter mice were generated and characterized previously [35] , [36] . MCMV clones were grown on M2-10B4 cells and partially purified as described [56] , with the following modification: upon ultracentrifugation , the virus pellet was resuspended in 1 . 2 ml of Virus standard Buffer ( 0 . 05 M Tris , 0 . 012 M KCl , 0 . 005 M EDTA ) and centrifuged in a microcentrifuge for 5 minutes at 3000×g . The clear supernatant was harvested , aliquoted and stored at −80°C . The BAC-derived wild-type MCMV ( MCMV WT ) [57] and MCMVr [23] have been described previously . The 230 kb MCMV BAC Δm157 eGFP was generated by homologous recombination of a linearized PCR fragment expressing the eGFP gene under the control of the minimal CMV promoter into the m157 genomic region of the pSM3fr BAC , essentially as described [58] . In brief , the gene was inserted by a two-step mutagenesis procedure , where in the first step the gene was introduced into the BAC , along with a kanamycin resistance gene ( kan ) flanked by frt sites , at nucleotide positions 216291 to 216874 , thus replacing most of the m157 gene , including its start codon . In a subsequent step , kan was excised by transient expression of flip recombinase , and recombined clones were selected by kanamycin sensitivity , thus generating the Δm157/eGFP pSM3fr plasmid . The recombinant virus MCMV IE1/3flox contains two loxP sequences which flank the open reading frames of the immediately early genes ie1 and ie3 and is a derivative of MCMV WT [57] . MCMV IE1/3flox was generated by two-step recombination mutagenesis using the galK selection system and modified to include antibiotic resistance selection in the first mutagenesis step [59] . A linear PCR-derived recombination fragment encoding galK and kanamycin resistance ( KanR ) was amplified from the pGPS/galKn plasmid [59] using primers P9 and P11 ( for primer and construct sequences see supplementary table S4 ) , inserted into SW102 E . Coli carrying the MCMV WT BAC genome and recombined BAC clones were selected on kanamycin plates . The synthetic DNA construct C1 ( Geneart ) was subsequently introduced , replacing the GalK/Kan gene with a loxP site at nucleotide position 177965–177974 according to the published MCMV genome annotation [60] . The second loxP-site was inserted with the same method , using primers P43 and P45 in the first mutagenesis step and the synthetic DNA product C2 ( Eurofins MWG Operon ) inserted in the second step at nucleotide position 182837–182846 . The entire sequence of the final BAC clone was sequenced in an Illumina sequencer to exclude illegitimate recombination events . The newly generated BACs were transferred into MEFs and reconstituted viruses grown as described above . VSV-GFP [61] and MHV68 GFP [62] were grown as described previously . Confluent monolayers of non-cycling LSECs or NIH3T3 were infected with MCMV WT or MCMVr at the Multiplicity of infection ( MOI ) of 0 . 1 . After 1 h , the cells were washed with PBS , supplied with fresh medium and incubated for 6 days . SN were harvested in triplicates and stored at −70°C until they were titrated on MEFs . Confluent monolayers of non-cycling LSECs were infected in 96-well plates with MCMVr , MHV68 GFP or VSV GFP at indicated MOIs . LSECs were treated with recombinant mouse IFNβ ( PBL Interferon Source , Piscataway , NJ ) as follows: ( 1 ) untreated LSECs ( −IFNβ ) were cultivated with normal medium throughout the experiment . ( 2 ) IFNβ-treated LSECs ( +IFNβ ) were stimulated 24 h before the infection and supplied with IFNβ throughout the experiment . ( 3 ) LSECs , in which the IFNβ was retracted ( +/−IFNβ ) were stimulated 24 h before the infection and cultivated for 7 days in the presence of IFNβ . At 7 dpi , the IFN-containing medium was exchanged with normal medium and cells were cultivated without IFNβ until the end of the experiment . For all conditions , the cells were supplied with fresh medium every 2–3 days . Infected cells were monitored by Fluorescence Microscopy for reporter gene expression at the indicated time points and wells that showed viral replication indicated by fluorescent cells , were classified as positive . LSECs were cultivated on chamber slides ( Thermo Scientific ) and stimulated for 24 h with IFNβ . The cells were stained with Daxx ( clone 25C12; 1∶25 ) rabbit mAb ( Cell Signaling ) according to the manufacture's protocol . In brief , cells were fixed with Formaldehyde , permeabilized with Triton-X-100 and anti-rabbit Alexa 488 ( clone B13C; 1∶200 ) was used as secondary antibody . The cells were mounted with VECTASHIELD ( Vector Laboratories ) prior to microscopic analysis . 6 to 10 weeks old C57BL/6 mice ( Janvier ) were intraperitoneally infected with 106 PFU of MCMVr and housed in SPF conditions throughout the experiment . Initial isolation of mouse liver non parenchymal cell ( NPC ) was performed according to a published protocol [63] . In brief , liver was perfused with 5 ml liver perfusion medium ( Gibco-Invitrogen , Paisley , UK ) and with 5 ml liver digestion medium ( Gibco-Invitrogen , Paisley , UK ) . Upon removal of the liver from the mouse , the liver was cut in small pieces , incubated for 30 min in liver digestion medium and gently pressed through a Nylon 100 µm cell strainer ( BD Falcon ) . Cells up from five livers were pooled , washed in PBS , resuspended in 40% Percoll ( Biochrom ) , gently overlaid onto 70% Percoll , and centrifuged at 750× g for 20 min . NPC collected from the interface were washed twice and resuspended in PBS/1%FCS . Upon red blood cell lysis , LSECs were isolated from NPCs by immunomagnetic sorting . For this purpose , cells were counted and resuspended in 10 µl of antimouse-CD146–conjugated magnetic beads ( Miltenyi Biotec ) and 90 µl of PBS+1% FCS per 107 nucleated cells , incubated for 15 minutes at 4°C and magnetically separated according to the manufacturer's protocol . Isolated LSECs were maintained in RPMI supplemented with 10% fetal bovine serum ( FBS ) ( PAN Biotech , Aidenbach Germany ) , penicillin ( 100 U/mL ) , streptomycin ( 100 µg/mL ) , L-glutamine ( 2 mM ) , 1 mM sodium pyruvate and 0 . 2 mM 2-mercaptoethanol ( Gibco-Invitrogen , Paisley , UK ) on plates coated with 0 . 5% gelatin ( Sigma , St . Louis , MO , USA ) . Cells were seeded and cultivated in an incubator at 37°C , 7% CO2 and 5% O2 , at maximal humidity . LSECs were treated with recombinant mouse IFNβ and monitored by fluorescent microscopy for signs of virus reactivation as detailed above . For plaque assay , LSECs were treated with recombinant mouse IFNβ ( PBL Interferon Source , Piscataway , NJ ) as described above with the following modifications: LSECs were treated with 100 U/mL and the IFNβ was removed after seven days of cultivation . Triplicate SN were stored at −70°C and titrated on IFNAR−/− MEFs . NIH3T3 IRF7-mCherry cells were stimulated with IFNβ ( 500 U/mL ) . The cells were trypsinized 24 h later , resuspended in PBS and sorted in mCherryhigh and mCherrylow populations using a FACS Aria II ( BD Bioscience ) cell sorter . For the stimulation of IFN-β+/Δβ-luc mice , poly I:C ( 100 µg/mouse ) was injected i . v . or the mice were mock injected with PBS only . To visualize the reporter gene in IFN-β+/Δβ-luc and Mx2-luc mice , the mice were injected i . v . with 150 mg/kg of D-luciferin in PBS ( Calipers ) , anesthetized using Isofluran ( Baxter ) and monitored using an IVIS 200 imaging system ( Calipers ) . Photon flux was quantified using the Living Image 3 . 0 software ( Calipers ) . Overlays were analyzed using the Living Image 4 . 1 software . Relative intensities of emitted light were presented as pseudocolor overlays ranging from red ( most intense ) to black ( least intense ) . Data were expressed as radiance , quantified as photons/sec/cm2/sr . Steradian ( sr ) refers to the photons emitted from a unit of solid angular measure . Cells were treated with IFNβ ( 500 U/ml ) for 24 h prior to infection . Cells were infected with WT MCMV at a nominal MOI of 1 . Virus was allowed to absorb for 5 minutes at 2000 rpm in a tissue culture centrifuge and was removed immediately thereafter , which increases the infectivity rate by a factor of 10 as compared to cells infected with the same amount of virus for 1 h , in the absence of centrifugation ( See Figure S5 ) . Importantly , this increased the time resolution to the 5 minutes of virus absorption . The labeling and isolation of nascent RNA was performed for 1 h as described [29] , and biological duplicates of the transcriptome ( 100 ng of nascent RNA per sample ) were used for TruSeq RNA Library construction using TruSeq RNA sample preparation kit ( Low-Throughput protocol ) according to manufacturer's protocol . The final amplified library was purified using AMPure XP Beads ( Agencourt ) . Quality of TruSeq Libraries were checked using Agilent Technologies 2100 Bioanalyzer and run on Genome Analyzer IIx ( Illumina Inc . ) in single end mode with length of 36 nt per read . The program BWA [64] was used to align the reads to a reference genome composed of the mouse genome ( version mm10 ) with the MCMV genome ( NC_004065 . 1 ) inserted as an extra chromosome . Reads were read into the R statistical language version 2 . 15 [65] counted , and evaluated with the R package edgeR [66] following the edgeR tutorial . Annotation for the mouse was downloaded using the GenomicFeatures R package ( available from the Bioconductor website ) from the UCSC database , while viral annotation was created from the NCBI Genbank NC_004065 GFF file using GenomicFeatures . Mouse and virus data were analysed both separately and together , and reads per kilobase per million ( RPKM ) values were generated . Significantly differently expressed genes were determined by edgeR using two replicates . shRNAs-sequences targeting murine DAXX , SP100 , and PML or non-coding ( NC ) shRNA were generated by using the online tool from Integrated DNA Technologies ( IDT ) or the database from the RNAi-consortium . Design of shRNA-vector inserts for of pRNA-U6/Neo ( GeneScript #SD1201 ) was performed according to the manufacturer's manual . Sense and antisense siRNA sequences were ordered as loop-sequences annealed before ligation into the shRNA-vector . Single clones were selected and sequenced . Both Sense and Antisense sequences of the used shRNAs are listed in Table S3 . LSECs were transfected with 2 µg plasmid DNA encoding for shRNA targeting Daxx , SP100 , PML or a non-targeting control-shRNA . 24 h following transfection , medium was exchanged by RPMI supplemented with IFNβ ( 500 U/mL ) , incubated for 24 h upon which the RNA was extracted from cells with TRIzol ( Invitrogen ) , according to the manufacturers protocol . cDNA was synthesised with SyperScript II ( Invitrogen ) and oligo ( dT ) 12–18 primers according to manufacturer's recommendation . qRT-PCR for genes of interest was performed using peqGOLD REAL-TIME ( Peqlab ) and SYBR Green in a LightCycler 480 ( Roche ) and the results were normalised to GAPDH . | Cytomegalovirus ( CMV ) is a widespread herpesvirus that establishes a détente with the host immune system . Therefore , the CMV reactivates from latency in immunocompromised hosts , resulting in life-threatening disease of the vulnerable patients . However , the exact mechanism by which the immune system keeps CMV at bay remains incompletely understood . To address this question , we have used a reporter system , based on infection of cells with the mouse CMV . Our results showed that interferon ( IFN ) , a well-known antiviral protein , blocks CMV replication at the earliest stages after the virus has entered the cell . More importantly , removing the IFN from the infected cells restarted MCMV replication , indicating that its effects are consistent with viral latency . We showed that IFN blocked virus replication by inducing the expression of proteins located in the nuclear domain 10 ( ND10 ) , a compartment in the nucleus of cells to which the incoming viral genomes are directed . Similarly , IFN was sufficient to block CMV reactivation from cells of latently infected mice . In conclusion , IFN had the ability to drive CMV into a quiescent state matching the formal definition of latency and was sufficient to prevent reactivation of bona fide latent CMV . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"medicine",
"viral",
"latency",
"infectious",
"diseases",
"immunity",
"virology",
"innate",
"immunity",
"immunology",
"cytomegalovirus",
"infection",
"biology",
"microbiology",
"viral",
"diseases",
"viral",
"replication"
] | 2014 | Reversible Silencing of Cytomegalovirus Genomes by Type I Interferon Governs Virus Latency |
As the most widespread tick-borne arbovirus causing infections in numerous countries in Asia , Africa and Europe , Crimean-Congo Hemorrhagic Fever Virus ( CCHFV , family Nairoviridae ) was included in the WHO priority list of emerging pathogens needing urgent Research & Development attention . To ensure preparedness for potential future outbreak scenarios , reliable diagnostic tools for identification of acute cases as well as for performance of seroprevalence studies are necessary . Here , the CCHFV ortholog of the major bunyavirus antigen , the nucleoprotein ( NP ) , was recombinantly expressed in E . coli , purified and directly labeled with horseradish peroxidase ( HRP ) . Employing this antigen , two serological tests , a μ-capture ELISA for the detection of CCHFV-specific IgM antibodies ( BLACKBOX CCHFV IgM ) and an IgG immune complex ( IC ) ELISA for the detection of CCHFV-specific IgG antibodies ( BLACKBOX CCHFV IgG ) , were developed . Test performance was evaluated and compared with both in-house gold standard testing by IgM/IgG indirect immunofluorescence ( IIF ) and commercially available ELISA tests ( VectoCrimean-CHF-IgM/IgG , Vector-Best , Russia ) using a serum panel comprising paired samples collected in Kosovo during the years 2013–2016 from 15 patients with an acute , RT-PCR-confirmed CCHFV infection , and 12 follow-up sera of the same patients collected approximately one year after having overcome the infection . Reliably detecting IgM antibodies in all acute phase sera collected later than day 4 after onset of symptoms , both IgM ELISAs displayed excellent diagnostic and analytical sensitivity ( 100% , 95% confidence interval ( CI ) : 85 . 2%–100 . 0% ) . While both IgG ELISAs readily detected the high IgG titers present in convalescent patients approximately one year after having overcome the infection ( sensitivity 100% , 95% CI: 73 . 5%–100 . 0% ) , the newly developed BLACKBOX CCHFV IgG ELISA was superior to the commercial IgG ELISA in detecting the rising IgG titers during the acute phase of the disease . While all samples collected between day 11 and 19 after onset of symptoms tested positive in both the in-house gold standard IIFT and the BLACKBOX CCHFV IgG ELISA ( sensitivity 100% , 95% CI: 71 . 5%–100 . 0% ) , only 27% ( 95% CI: 6 . 0%–61 . 0% ) of those samples were tested positive in the commercial IgG ELISA . No false positive signals were observed in either IgM/IgG ELISA when analyzing a priori CCHFV IgM/IgG negative serum samples from healthy blood donors , malaria patients and flavivirus infected patients as well as CCHFV IgM/IgG IIFT negative serum samples from healthy Kosovar blood donors ( for BLACKBOX CCHFV IgM/IgG: n = 218 , 100% specificity , 95% CI: 98 . 3%–100 . 0% , for VectoCrimean-CHF-IgM/IgG: n = 113 , 100% specificity , 95% CI: 96 . 8%–100 . 0% ) .
Being endemic in a variety of countries in Asia , Africa , the Middle East and Southeastern Europe [1] , the Crimean-Congo Hemorrhagic Fever Virus ( CCHFV , family: Nairoviridae , genus: Orthonairovirus; [2] ) is the geographically most widespread tick-borne pathogen [3 , 4] . An infection with this zoonotic virus occurs either by tick bite or contact with body fluids or tissues of viremic humans or animals [5] . After an incubation time of 1–9 days , unspecific flu-like symptoms develop during the first week of disease including high fever , myalgia , arthralgia and headache [5] . Following this prehemorrhagic phase , the disease may progress to the hemorrhagic phase manifesting in petechiae , hematomas , gastrointestinal bleeding ( hematemesis , melena ) , urinary tract bleeding ( hematuria ) and/or respiratory tract bleeding ( hemoptysis ) [6] . Case fatality rates ( CFRs ) range widely between 5% and more than 50% [6] with decreased platelet counts , elevated aspartate transferase and alanine transferase and decreased fribrinogen levels being predictive of fatal outcome [7] . In addition , no CCHFV-specific humoral immune response is observable in the majority of fatal cases [7 , 8] . CCHF affects residents of endemic countries but can also be imported into non-endemic areas by travelers [1] . Furthermore , the geographic distribution of the disease is expanding as evidenced by the prevalence of CCHFV-infected ticks [9] and the recent occurrence of autochthonous CCHF in Spain [10] . Up to now , neither a vaccine with the potential to gain widespread international regulatory approval nor a specific treatment is available [4] , which restricts prophylaxis to exposure prevention and therapy to mainly supportive measures . Due to the severity of the disease , the lack of specific prophylactic and therapeutic options and its epidemic potential , CCHFV was included in the 2015 WHO priority list of diseases needing urgent R&D attention [11] , in particular the development and improvement of diagnostic tools . Although the specificity of commercially available serological assays for the detection of CCHFV-specific antibodies in human sera was found to be excellent in a collaborative evaluation of CCHF diagnostic tests [3] , the reported assay sensitivities of 87 . 8% ( range 75 . 2%–95 . 3% ) for IgM ELISA testing and 80% ( range 66 . 9%–90 . 2% ) for IgG ELISA testing [3] left room for improvements . Therefore , we developed and validated both a CCHFV μ-capture ELISA ( BLACKBOX CCHFV IgM ELISA ) and a CCHFV IgG immune complex ( IC ) ELISA ( BLACKBOX CCHFV IgG ELISA ) based on the directly horseradish-peroxidase ( HRP ) labeled , recombinantly produced CCHFV ortholog of the major bunyavirus antigen [12] , the nucleoprotein ( NP ) . Assay performance was evaluated in comparison with in-house gold standard testing by IgM/IgG indirect immunofluorescence ( IIF ) and commercially available IgM/IgG ELISA tests ( VectoCrimean-CHF-IgM/IgG , Vector-Best , Russia ) using a serum panel comprising paired samples from 15 Kosovar patients with an RT-PCR-confirmed acute CCHFV-infection and 12 follow-up samples from the same group of patients taken approximately one year after they had overcome the disease .
The plasmid pOPINJ-CCHFV-NP was kindly provided by Sophia Reindl ( Department of Virology , BNITM ) . Briefly , a cDNA fragment encoding the nucleoprotein ( NP ) of CCHFV strain Afg09-2990 ( ADQ57288 , [13] ) was cloned into the prokaryotic expression vector pOPINJ [14] . The resulting plasmid pOPINJ-CCHFV-NP encodes a 82 kDa fusion protein comprising an N-terminal 6 x His-Tag followed by Glutathione-S-transferase ( GST; 26 kDa ) , a 3C protease cleavage site and the CCHFV-NP ( 54 kDa ) . Expression in E . coli: The plasmid pOPINJ-CCHFV-NP was transformed into E . coli pAPlacIQ cells . Cells were grown at 37°C to an optical density at 600 nm of 0 . 5–0 . 6 in Luria-Bertani ( LB ) medium supplemented with kanamycin and ampicillin , then expression of the CCHFV-NP fusion protein was induced by the addition of isopropyl-β-D-thiogalactosid ( IPTG ) . Bacteria were further incubated over night at 18°C and then harvested by centrifugation . Cell lysis: Cells were resuspended in lysis buffer ( 50 mM NaH2PO4 pH 8 . 0 , 300 mM NaCl , 10 mM imidazole ) supplemented with phenylmethylsulfonylfluorid ( PMSF ) and lysozyme and incubated at 4°C for 30 min before sonification . After sonification , DNAse was added to a final concentration of 10 μg/ml . The lysate was incubated for 15 min on ice and then centrifuged at 10000 x g for 20 min at 4°C . Ni-NTA chromatography ( native conditions ) : After centrifugation of the lysate , the resulting supernatant was incubated with pre-equilibrated Ni-NTA-agarose ( Qiagen ) for 1 h at 4°C . The matrix was transferred to a column and the flow-through was discarded . After washing the column , bound protein was eluted with Elution Buffer ( 50 mM NaH2PO4 pH 8 . 0 , 300 mM NaCl , 250 mM imidazole ) . On-column-cleavage of His6-GST tandem tag: The main fractions were pooled and the buffer was exchanged to Binding Buffer ( 50 mM NaH2PO4 pH 7 . 2 , 300 mM NaCl , 5 mM dithiothreitol ( DTT ) ) using Zeba Spin Desalting Columns ( Thermo Scientific ) . Subsequently , tagged recombinant CCHFV-NP was coupled to pre-equilibrated Glutathione HiCap matrix ( Qiagen ) . Unbound material was removed from the column by washing with Binding Buffer . Recombinant GST-tagged 3C-protease was homogenously added to the column matrix and cleavage was performed over night at 4°C with intermittent gentle shaking . Subsequently , CCHFV-NP was eluted from the column and concentrated using Amicon Ultra 15 micro concentrators ( Merck Millipore , cut-off: 10k ) . Size exclusion chromatography ( SEC ) : The eluted , concentrated protein was further purified by SEC using a Superdex 200 10/300 GL-column on an ÄKTA pure chromatography system ( GE Healthcare ) . Peak fractions containing recombinant CCHFV-NP were pooled and concentrated using Amicon Ultra 15 micro concentrators ( cut-off: 10k ) . HRP-labeling: For labeling recombinant CCHFV-NP with HRP , HRP ( Sigma ) was activated for coupling with sodium periodate ( final concentration 3 . 5 mg/ml ) for 20 min at room temperature ( RT ) . Subsequently , dialysis against 1 mM acetate buffer pH 4 . 4 was performed at 4°C over night . Activated HRP was incubated with recombinant CCHFV-NP for 4 h at 4°C and sodium borhydride was added at a final concentration of 0 . 2 mg/ml . After a 5 min incubation at RT , the labeling reaction mix was diluted 1:10 with Freezing Buffer ( 0 . 5 x phosphate-buffered saline ( PBS ) , 1% bovine serum albumin ( BSA ) , 0 . 5% fetal calf serum ( FCS ) , 1% Nonidet P40 , 50% glycerol ) and stored at - 20°C . IgM μ-capture ELISA ( BLACKBOX CCHFV IgM ) : ELISA plates ( MaxiSorp , Nunc ) were coated over night at 4°C with 4 μg/ml goat anti-human IgM ( KPL ) in PBS pH 7 . 4 . After washing the plates two times with Blocking Buffer ( PBS pH 7 . 4 , 0 . 25% BSA , 0 . 05% Tween 20 ) , plates were blocked with Blocking Buffer for 2 h at RT . Plates were washed two times with PBS pH 7 . 4 , stabilized and dried using a commercial plate stabilizer ( Liquid PlateSealer , Candor ) according to the manufacturer’s recommendations . For detection of anti-CCHFV-NP IgM antibodies in human sera , serum samples were diluted 1:100 in Serum Dilution Buffer ( PBS pH 7 . 4 , 0 . 05% Proclin 300 , 0 . 01% phenol red ) . Per well , 50 μl of diluted serum was applied and the plates were incubated for 60 min at RT ( 23°C ) . After washing the plates three times with Wash Buffer ( 100 mM Tris/HCl pH 7 . 4 , 150 mM NaCl , 0 . 05% Tween 20 , 0 . 005% ProClin 300 ) , 300 μl/well ) , 50 μl HRP-labeled CCHFV-NP ( final dilution 1:50 , 000 in Conjugate Dilution Buffer ( PBS pH 7 . 4 , 1% BSA , 0 . 5% FCS , 1% Nonidet P40 , 0 . 1% ProClin 300 ) ) was added per well and the plates were incubated for 60 min at RT ( 23°C ) . Plates were washed again three times with Wash Buffer ( 300 μl/well ) , and 100 μl SureBlue Reserve TMB Microwell Peroxidase Substrate ( KPL ) was added per well . Plates were incubated for 20 min at RT ( 23°C ) and the reaction was then stopped by the addition of 100 μl 1N sulfuric acid ( Merck ) per well . The HRP reaction product was quantified by measuring optical density at 450 nm and 620 nm on a Spectrostar Nano ELISA reader ( BMG Labtech ) . IgG Immune-Complex ( IC ) ELISA ( BLACKBOX CCHFV IgG ) : The extracellular part of human CD32 ( FcγRIIa H131 ) was recombinantly expressed in E . coli and purified basically as described previously [15] . ELISA plates ( MaxiSorp , Nunc ) were coated at 4°C with 5 μg/ml recombinant CD32 protein in Coating Buffer ( PBS pH 7 . 4 , 0 . 01% NaN3 , 0 . 01% phenol red ) . After washing three times with 100 mM Tris pH 7 . 4 , 150 mM NaCl , plates were stabilized and dried using a commercial plate stabilizer ( Liquid PlateSealer , Candor ) according to the manufacturer’s recommendations . For detection of anti-CCHFV-NP antibodies in human sera , serum samples were diluted 1:50 in Serum Dilution Buffer ( PBS pH 7 . 4 , 0 . 05% Proclin 300 , 0 . 01% phenol red ) . Per well , 25 μl of diluted human serum were mixed with 25 μl HRP-labeled CCHFV-NP diluted 1:125 , 000 in Conjugate Dilution Buffer ( PBS pH 7 . 4 , 1% BSA , 0 . 5% FCS , 1% Nonidet P40 , 0 . 1% ProClin 300 ) ; final concentrations in well: serum 1:100 , HRP-labeled CCHFV-NP 1:250 , 000 . Plates were sealed with adhesive foil and incubated for 24 h at 4°C . Plates were washed three times with Wash Buffer ( 100 mM Tris/HCl pH 7 . 4 , 150 mM NaCl , 0 . 05% Tween 20 , 0 . 005% ProClin 300; 300 μl/well ) ) and 100 μl SureBlue Reserve TMB Microwell Peroxidase Substrate ( KPL ) was added per well . Plates were incubated for 20 min at RT ( 23°C ) and the reaction was then stopped by the addition of 100 μl 1 N sulfuric acid ( Merck ) per well . The HRP reaction product was quantified by measuring optical density at 450 nm and 620 nm on a Spectrostar Nano ELISA reader ( BMG Labtech ) . The IgG immune complex technology employing recombinant CD32 is protected by European ( EP2492689 ) and international ( CN103460048 , HK1192320 , CA2823107 , US2014080120 ) patents owned by the Bernhard Nocht Institute for Tropical Medicine ( BNITM ) . VectoCrimean-CHF-IgM/IgG ELISA ( Vector-Best , Koltsovo , Russia ) : assays ( IgM: μ-capture ELISA , IgG: indirect ELISA using immobilized CCHFV antigen ) were performed and evaluated according to the manufacturer’s instructions . Statistical analysis ( One-way ANOVA/Tukey’s multiple comparison test , Fisher’s exact test , calculation of 95% confidence intervals ) was performed with GraphPad Prism . Receiver operating characteristic ( ROC ) curves were generated using MedCalc . S segment nucleotide sequences of CCHFV strains were downloaded from the National Center for Biotechnology Information ( NCBI ) Nucleotide database . Alignments of derived protein sequences and generation of identity/similarity matrices were performed with the ClustalW alignment tool of the MacVector ( version 12 . 7 . 5 ) software . CCHF patient sera: Paired serum samples were obtained from 15 Kosovar patients with a PCR-confirmed CCHFV-infection ( 13/15 ( 87% ) male; 2/15 ( 13% ) female; median age: 40 years ( range 10–75 ) between 2013 and 2015 ( 2013: n = 10 , 2014: n = 4 , 2015: n = 1 ) in the course of the joint research project ( BNITM , University of “Hasan Prishtina” and National Institute of Public Health , Pristhina , Kosovo ) “Diagnosis and surveillance of Crimean-Congo hemorrhagic fever ( CCHF ) in Kosovo” , funded by the German Ministry of Foreign Affairs ( Project-No 68777 EN 02761868 ) . Eight of 15 ( 53% ) patients had a known history of tick bite ( onset of symptoms after tick bite median 2 . 5 days ( range 1–5 days ) ) . From each of the 15 patients , one “early” sample collected between day 2 and day 14 after onset of symptoms ( median: 5 days ) and one “late” sample collected 3 to 23 days after the “early” sample ( median: 10 days ) between day 8 and day 36 after onset of symptoms ( median: 17 days ) were analyzed . CCHFV-infection has been proven by RT-PCR ( RealStar CCHFV RT PCR Kit , Altona Diagnostics ) for all 15 patients in a serum sample collected directly after admission to the hospital . Nevertheless , for five patients this sample was not available for ELISA testing due to scarcity of material . Thus , for all patients , the earliest available serum sample ( which may or may not be PCR positive ) was chosen as “early” sample for the ELISA analysis . In addition , “convalescent” samples were obtained from 12 patients approximately one year after recovery from CCHFV infection . IgM/IgG status of patient samples was characterized by in-house IgM/IgG IIFT using acetone-fixed Vero cells infected with CCHFV strain ArD39554 as described previously [16] . Sera were inactivated by the addition of Triton X-100 to a final concentration of 1% prior to serological testing . Sera from healthy blood donors from Kosovo: Sera were obtained from 98 healthy donors ( 77/98 ( 79% ) male; 21/98 ( 21% ) female; median age: 33 years ( range 20–60 ) ) . All sera had previously been tested negative by in-house CCHFV IgM/IgG IIFT . DENV: 8 DENV IgM/IgG positive sera were obtained from European travelers with a PCR confirmed DENV infection ( 2 x DENV1 , 2 x DENV2 , 2 x DENV3 , 2 x DENV4 ) . TBEV: 2 TBEV IgM/IgG positive sera were obtained from Biomex ( Germany ) . Malaria: 27 sera from patients with an acute P . falciparum malaria ( confirmed by IIFT ) and 6 sera from patients with an acute P . malariae infection ( confirmed by IIFT ) were obtained from the BNITM diagnostics department ( section parasitology ) . Healthy blood donors ( HD ) : HD Europe: 49 sera from healthy blood donors from Europe were taken at the BNITM between the years 2012 and 2016 . HD Asia/HD South America: Sera of healthy blood donors from Lao PDR ( n = 14 ) and Colombia ( n = 14 ) were collected during the projects “Savannakhet-Hamburg Research Program on Neglected Diseases” and “Valledupar-Hamburg research programme on diagnostics and research on tropical and emerging infections” in 2012 and 2013 , respectively . Healthy blood donors from Colombia had been vaccinated against YF . The study complies with the Declaration of Helsinki . Written informed consent was obtained from all individuals or , in case of minors , from parents or legal guardians before enrollment . Data privacy protection was guaranteed by anonymization of serum samples . Collection of serum samples was approved by the Ethics Committee of the University of Prishtina “Hasan Pristhina” ( sera from CCHF patients and healthy blood donors from Kosovo ) , the Ethics Committee of the Lao People’s Democratic Republic ( sera from healthy blood donors from Lao PDR , no . 030/NECHR ) , the Ethics Committee of the Hospital Rosario Pumarejo de Lopez of Valledupar/Colombia ( sera from healthy blood donors from Colombia ) and the Ethics Committee of the Ärztekammer Hamburg ( sera from Dengue fever patients , malaria patients and healthy blood donors from EU , no . PV4608 ) .
In order to generate highly purified recombinant CCHFV-NP for use as an antigen in CCHFV ELISA applications a fusion protein consisting of an N-terminal His/GST-tandem tag and the full length CCHFV-NP separated by a 3C protease cleavage site ( Fig 1A ) was expressed in E . coli ( Fig 1B ) . The recombinant fusion protein was purified from the soluble fraction of the bacterial lysate by Ni-NTA affinity chromatography ( Fig 1C ) . After removal of the tandem tag with 3C protease by on-column cleavage on glutathione matrix ( Fig 1D ) , CCHFV-NP was further purified by size exclusion chromatography and directly labeled with horseradish-peroxidase ( HRP ) . Using the HRP-labeled recombinant CCHFV-NP ( CCHFV-NPHRP ) as antigen , both a μ-capture ELISA protocol for the detection of human anti-CCHFV IgM antibodies ( designated as “BLACKBOX CCHFV IgM” ) and an IgG IC ELISA protocol for the detection of human anti-CCHFV IgG antibodies ( designated as “BLACKBOX CCHFV IgG” ) were developed . For the BLACKBOX CCHFV IgM ELISA , plates were coated with anti-human IgM antibodies and subsequently incubated for 1 h at RT with human serum and CCHFV-NPHRP . After washing , binding was quantified by measuring optical density after incubation with a chromogenic HRP-substrate . For the BLACKBOX CCHFV IgG ELISA , plates were coated with a recombinant fragment of the human FcγR CD32 [15 , 17] . Human serum and CCHFV-NPHRP were co-incubated for 24 h at 4°C on the plates . After washing , binding of immune complexes was quantified by measuring optical density after incubation with a chromogenic HRP-substrate . Titration of an IgM/IgG IIFT positive patient serum and analysis of a panel of a priori CCHFV IgM/IgG negative sera from healthy human blood donors and malaria patients indicated a broad detection range and low background signal of the assays ( Fig 2 ) . Furthermore , intra- and inter-assay variation ( Table 1 , Table 2 ) as well as inter-laboratory variation ( Table 3 , Table 4 ) were analyzed , revealing a high reproducibility of both assay systems ( mean intra-assay CV for positive sera < 5%; mean inter-assay CV for positive sera < 10% , inter-laboratory CV for positive serum < 10% ) . For assay validation , a serum panel comprising serum samples from 15 CCHF patients from Kosovo ( Table 5 ) was analyzed with the BLACKBOX CCHFV IgM ELISA ( Fig 3A ) and the BLACKBOX CCHFV IgG ELISA ( Fig 3B ) . From each of the 15 patients , one “early” sample collected between day 2 and day 14 after onset of symptoms ( median: 5 days ) and one “late” sample collected 3 to 23 days after the “early” sample ( median: 10 days ) between day 8 and day 36 after onset of symptoms ( median: 17 days ) were tested . In addition , “convalescent” samples taken from 12 patients approximately one year after recovery from CCHFV infection were analyzed . To evaluate assay specificity , 120 a priori CCHFV IgM/IgG negative sera originating from either healthy blood donors from Germany ( Europe ) , Colombia ( South America ) and Lao PDR ( Asia ) ( n = 77 ) , malaria patients ( n = 33 ) or flavivirus infected patients ( n = 10 ) were tested along with 98 CCHFV IgM/IgG IIFT negative sera from healthy blood donors from Kosovo ( Fig 3A , Fig 3B , panels A and B in S1 Fig ) . For determination of the optimal assay cut-offs , ROC analysis was performed using specific subsets of CCHF patient sera corresponding to the respective assay’s intended diagnostic purpose ( S2 Fig ) . The main intention of performing an IgM ELISA is the diagnosis of acute infections as early in the course of disease as possible , whereas IgG testing rather becomes relevant at later stages of acute disease progression and for seroprevalence studies on healthy individuals . Therefore , assay cut-offs were optimized for optimal differentiation of PCR negative “early” and “late” samples from the a priori negative control samples for the BLACKBOX CCHFV IgM ELISA ( panel A in S2 Fig , cut-off: 0 . 129 ) and PCR negative “late” and “convalescent” samples from the a priori negative control samples for the BLACKBOX CCHFV IgG ELISA ( panel B in S2 Fig , cut-off: 0 . 161 ) , respectively . When applying these cut-offs to the complete tested serum panel , both the BLACKBOX CCHFV IgM ELISA ( Fig 3A , Table 6 ) and the BLACKBOX CCHFV IgG ELISA ( Fig 3B , Table 7 ) display a 100% specificity ( 95% CI: 98 . 3%–100 . 0% ) and a 100% sensitivity ( 95% CI: 78 . 2%–100 . 0% ) when testing “late” serum samples . While all five PCR negative “early” samples tested positive in the BLACKBOX CCHFV IgM ELISA , CCHFV-specific IgM antibodies could only be detected in 40% of the PCR positive “early” samples ( Fig 3A , Table 6 ) . The BLACKBOX CCHFV IgG ELISA detected CCHFV-specific antibodies weakly in 2 out of 10 PCR positive “early” samples , but generated a clearly positive result in four out of five PCR negative “early” samples ( Fig 3B , Table 7 ) . For all 6 patients whose PCR positive “early” samples tested negative in both the BLACKBOX CCHFV IgM ELISA and the BLACKBOX CCHFV IgG ELISA , clear IgM and IgG seroconversion was observed from the “early” to the “late” sample ( panels A and B in S3 Fig ) . Surprisingly , 9 out of 12 “convalescent” serum samples taken from CCHF patients more than one year after having overcome the disease tested weakly positive in the BLACKBOX IgM ELISA ( Fig 3A , Table 6 ) . Nevertheless , a clear differentiation between an acute CCHFV infection and the convalescent state is possible by calculating the IgG/IgM ratio defined as the quotient of the optical densities obtained when performing the BLACKBOX CCHFV IgG ELISA and the BLACKBOX CCHFV IgM ELISA , respectively ( Fig 4A ) . While the IgG/IgM ratio in the patient samples taken during the acute phase of the disease ranges from 0 to < 3 . 0 , the “convalescent” patient samples display a significantly higher IgG/IgM ratio ( Fig 4A ) . All CCHF patient samples had previously been analyzed with CCHFV IgM and CCHFV IgG IIFT ( Table 5 ) . While the BLACKBOX CCHFV IgM ELISA displayed a comparable sensitivity as the IgM IIFT for all sample types tested ( Table 6 ) , the CCHFV IgG IIFT detected low IgG titers in 5 of the 10 PCR positive “early” samples while only 2 of those samples were recognized as IgG positive by the BLACKBOX CCHFV IgG ELISA ( Table 7 ) . Nevertheless , 4 of 5 PCR negative “early” samples as well as all “late” ( 15/15 ) and “convalescent” ( 12/12 ) samples tested positive in both the CCHFV IgG IIFT and the BLACKBOX CCHFV IgG ELISA ( Table 7 ) . Thus , the performance of both ELISAs is comparable to gold standard testing when applied according to their respective intended diagnostic purpose ( BLACKBOX CCHFV IgM ELISA for acute diagnostics , BLACKBOX CCHFV IgG ELISA for later stages of disease progression and seroprevalence studies ) . The same CCHFV serum panel which had been analyzed with the BLACKBOX CCHFV IgM ELISA and the BLACKBOX CCHFV IgG ELISA ( 15 serum pairs from CCHF patients , 12 sera from convalescent patients collected approximately one year after the infection ) , 15 a priori CCHFV-IgM/IgG negative control sera and 98 CCHFV IgM/IgG IIFT negative sera from healthy Kosovar blood donors were tested with both the VectoCrimean-CHF-IgM ELISA ( Fig 3C , Fig 4B , panel C in S1 Fig , S2 Fig , S3 Fig ) and the VectoCrimean-CHF-IgG ELISA ( Fig 3D , Fig 4B , panel D in S1 Fig , S2 Fig , S3 Fig ) . While the BLACKBOX CCHFV IgM ELISA and the VectoCrimean-CHF-IgM ELISA gave identical results ( Table 6 ) , the BLACKBOX CCHFV IgG ELISA was found to be more efficient than the VectoCrimean-CHF-IgG ELISA ( Table 7 ) in detecting IgG antibodies in the “late” serum samples collected between day 8 and day 36 ( median: day 17 ) after onset of symptoms . To dissect this observation more thoroughly , the 30 paired “early” and “late” serum samples were classified according to the day after onset on which they were collected ( Fig 5 ) . While both the BLACKBOX CCHFV IgM ELISA ( Fig 5A , Table 8 ) and the VectoCrimean-CHF-IgM ELISA ( Fig 5C , Table 8 ) detected 100% of samples that were collected later than day 4 after onset of symptoms as positive ( 95% CI: 85 . 2%–100 . 0% ) , the BLACKBOX CCHFV IgG ELISA was significantly more sensitive than the VectoCrimean-CHF-IgG ELISA in detecting IgG antibodies in samples collected between day 11 and 19 after onset of symptoms ( Fig 5B , Fig 5D , Table 9; BLACKBOX CCHFV IgG: 100% ( 95% CI: 71 . 5%–100 . 0%; VectoCrimean-CHF-IgG: 27% ( 95% CI: 6 . 0%–61 . 0% ) ) . Evaluation of the correlation of optical densities measured using the BLACKBOX and Vector-Best IgM/IgG tests and the antibody titers obtained by in-house CCHFV IgM/IgG IIFT ( Fig 6 ) revealed a marked difference between serum samples collected during the acute/early convalescent phase of the disease and follow-up samples collected approximately one year after the patients having overcome the disease . All follow-up samples generated a highly positive signal in the BLACKBOX CCHFV IgG test even a titer of 1:640 , the lowest CCHFV IgG IIFT titer observed in these samples . On the other hand , optical densities measured for acute/early convalescent phase sera with the identical CCHFV IgG IIFT titer varied strongly from sample to sample ( Fig 6B ) . Thereby , later sampling correlated with higher signals ( Fig 6B ) .
In this work , two ELISA tests , a CCHFV μ-capture ELISA designated as “BLACKBOX CCHFV IgM ELISA” and a CCHFV IgG Immune Complex ( IC ) ELISA termed “BLACKBOX CCHFV IgG ELISA” , were developed employing the CCHFV ortholog of the major bunyavirus antigen , the nucleoprotein NP [12] . In the native virus particle , NP associates with the viral genomic RNA to form ribonucleoprotein ( RNP ) complexes , it is involved in virus assembly and mediates crucial interactions with host cell components like the cytoskeleton and the translation machinery [18] . Structural analysis of the 54 kDa CCHFV NP [18–20] revealed a racket-like structure with two distinct domains ( “head” and “stalk” ) . Thereby , the stalk domain is formed by the central part of the 482 aa CCHFV NP polypeptide chain ( ≈ aa 180–300 ) and harbors the main antigenic region of the protein which is highly conserved between CCHFV strains from various endemic areas [21–23] . CCHFV NP has been used previously as an antigen in a variety of ELISA applications , either utilizing a combination of inactivated native virus lysate and an NP-specific monoclonal antibody [24] or purified recombinant NP produced either in insect cells [23 , 25 , 26] , mammalian cells [27] , E . coli [28 , 29] or even plants [30] . In these cases , most IgG ELISA applications were based on the concept of indirect ELISA [23 , 25–30] , whereas IgM ELISA applications were set up using μ-capture protocols [24 , 27] . In contrast , the BLACKBOX CCHFV IgG ELISA developed in this work employs the IgG immune complex ( IC ) binding principle . In this type of ELISA , originally rheumatoid factor ( RF ) was used as a capture molecule to bind immune complexes formed between pathogen-specific IgG antibodies and either native or recombinant viral antigens [16 , 24 , 31–33] . Recently , we showed that the method could be further improved by replacing RF by the recombinantly produced tandem immunoglobulin-like domain of the human FcγR CD32 [17] and we have already demonstrated the potential of this by now patented assay format for the sensitive and specific detection of DENV-subtype-specific IgG antibodies [15] . In nature , CCHFV circulates in an enzootic tick-vertebrate-tick cycle and can be transmitted from livestock to human by contact with blood or other body fluids during activities like slaughtering [34] . Thus , ELISA tests for animal sera like the double-antigen sandwich ELISA recently developed by Sas et al . [35] are needed for surveillance purposes . The pronounced interspecies cross-reactivity of the recombinant CD32 fragment [17] may facilitate the development of such tests . To evaluate the performance of the newly developed BLACKBOX CCHFV IgM/IgG ELISA tests , acute and convalescent phase sera from 15 patients with a PCR-confirmed CCHFV infection collected in Kosovo between 2013 and 2016 have been analyzed . Kosovo is a landlocked , southeastern European country on the central Balkan Peninsula with a population of 1 . 9 million . Due to a hot , dry climate favoring vector prevalence and extensive agricultural use , CCHFV is endemic in approximately half of its territory of 10 , 908 km2 [36] . Between 1954 and 2014 , 304 confirmed cases of CCHF were reported with a case fatality rate ( CFR ) of 21% with the highest incidence in late spring/early summer ( May–July ) [37] . A seroprevalence of up to 9% was found in inhabitants of hyperendemic areas , the majority ( 70% ) of seropositive individuals being males of advanced age ( median age 62 ) who had probably been exposed to the virus during professional activities like farming or slaughtering [36] . Correspondingly , a high seroprevalence of up to 20% was observed in livestock ( cows , goats , sheep ) being kept in endemic areas [36] . Recently , analysis of tick species revealed Hyalomma marginatum as the dominant species ( 90% of all collected specimen ) in endemic municipalities [38] . The percentage of infected H . marginatum ticks in endemic municipalities varied strongly between sampling sites ( range 1 . 25%–29 . 4% , average 11% ) [38] . Partial S segment sequences from all CCHFV-positive ticks could be classified as two different lineages , Kosovo I and II , in CCHFV clades VI ( Europe 2 ) and V ( Europe 1 ) , respectively . Thereby , sequences isolated from H . marginatum ticks cluster with Kosovar human-derived sequences in clade V ( Europe 1 ) , suggesting that the Kosovo outbreaks are most likely due to H . marginatum transmitted CCHFV [38] . Results obtained with the BLACKBOX IgM/IgG ELISAs were compared with the respective in-house serological gold standard ( CCHFV IgM/IgG IIFT ) and commercial ELISA tests ( VectoCrimean-CHF-IgM/IgG , Vector-Best , Russia ) that have been extensively used for seroprevalence studies in various countries ( e . g . Turkey [39–42] , Greece [43–47] , Bulgaria [48] , Afghanistan [49] , Kosovo [36] , Tunisia [50] , Ghana [51] ) and for the diagnosis of acute cases [52 , 53] . Recently , a collaborative evaluation of both the VectoCrimean-CHF-IgG and the VectoCrimean-CHF-IgM test in five reference laboratories revealed a sensitivity of 87 . 8% ( range 75 . 2%–95 . 3% ) for the VectoCrimean-CHF-IgM test and a sensitivity of 80 . 4% ( range 66 . 9%–90 . 2% ) for the VectoCrimean-CHF-IgG test when compared with the respective in-house reference serological tests [3] . Specificities were found to be 98 . 9% ( range 93 . 9%– 100 . 0% ) for the VectoCrimean-CHF-IgM tests and 100% ( range 95 . 8%–100% ) for the VectoCrimean-CHF-IgG test [3] . Here , we found that both the newly developed BLACKBOX CCHFV IgM ELISA ( employing recombinant CCHFV NP ) and the VectoCrimean-CHF-IgM test readily detected CCHFV specific IgM antibodies in all samples collected after day 4 after onset of symptoms corresponding to a sensitivity of 100% when compared to the in-house gold standard CCHFV IgM IIFT . Both the newly developed BLACKBOX CCHFV IgG ELISA employing recombinant CCHFV NP and the VectoCrimean-CHF-IgG test generated strong positive results with serum samples originating from convalescent patients approximately one year after having overcome the disease . However , the BLACKBOX CCHFV IgG ELISA was found to show a superior sensitivity in the early detection of the rising IgG antibody titer in acutely ill patients: the BLACKBOX CCHFV IgG ELISA recognized , like the in-house gold standard IgG IIFT , all tested patient samples which were collected after day 11 after onset of symptoms as IgG positive , whereas only 43% of those samples tested positive in the VectoCrimean-CHF-IgG test . This increased sensitivity of the BLACKBOX CCHFV IgG ELISA may also be of importance for the main intended use of an IgG assay , that is seroprevalence studies . When analyzing the correlation between CCHFV IgG IIFT titers and the OD values obtained using the BLACKBOX CCHFV IgG ELISA , a striking difference between serum samples collected during the acute/early convalescent phase of the disease and follow-up samples collected approximately one year after the patients having overcome the disease was observed . While the latter samples stably generated high OD values in the ELISA even at a rather low CCHFV IgG IIFT titer of 1:640 , the signal heights of different samples with an identical CCHFV IgG IIFT titer varied strongly for acute/early convalescent phase samples . Thereby , higher signals correlated with later sampling . Thus , the very first IgG antibodies generated during acute CCHFV infection presumably do not predominantly recognize the CCHFV nucleoprotein but rather other antigens like the virus envelope glycoprotein Gc and therefore are not detectable with serological tests employing solely CCHFV NP as an antigen . A similar observation was made by Putkuri et al . [54] for a member of orthobunyavirus genus , the Inkoo virus ( INKV ) . Here , the fraction of IgG antibodies recognizing either the nucleoprotein or the glycoprotein Gc , respectively , was found to differ considerably between serum samples from individuals with a long-standing immunity and samples collected during acute infection [54] . Both BLACKBOX CCHFV ELISAs employ the nucleoprotein of CCHFV strain Afg09-2990 ( belonging to S segment lineage IV ( Asia1 ) , [55] ) as antigen and were validated using CCHF serum samples originating from Kosovo . Thus , the patients’ infections were most likely caused by CCHFV strains belonging to S segment lineage V ( Europe 1 ) [38 , 56] . To further assess the potential of the two BLACKBOX tests to detect anti-NP-antibodies induced by infections with CCHFV strains belonging to different S segment lineages , a sequence alignment of the nucleoprotein ( NP ) amino acid ( aa ) sequences of six CCHFV strains belonging to five different S segment lineages [55] was performed ( panels A and B in S4 Fig ) . Sequence identities between the full length NP of the Afg09-2990 strain ( lineage IV , Asia 1 ) and strains ArD157856 ( lineage I , Africa 1 ) , UG3010 ( lineage II , Africa 2 ) , ArD39554 ( lineage III , Africa 3 ) , Kosovo Hoti ( lineage V , Europe 1 ) and Drosdov ( lineage V ( Europe 1 ) ) were found to be 94 . 8% , 96 . 9% , 97 . 3% , 96 . 9% and 96 . 7% , respectively ( panel C in S4 Fig ) . When the analysis was restricted to the NP stalk domain ( aa 180–300 ) harboring the main antigenic region of the protein [21–23] , even higher percentages of identity were observed with strains UG3010 ( 97 . 5% ) , ArD39554 ( 99 . 2% ) , Kosovo Hoti ( 99 . 2% ) , and Drosdov ( 99 . 2% ) ( panel C in S4 Fig ) . Therefore , the BLACKBOX tests most probably allow reliable detection of anti-CCHFV NP antibodies induced by CCHFV strains from different endemic regions of the world . This assumption is further supported by recent findings of Sas et al . [35] who demonstrated cross-detection of anti-CCHFV-NP antibodies in sera of animals infected with CCHFV strains belonging to S segment lineages III ( Africa 3 ) , IV ( Asia 1 ) , and V ( Europe 1 ) in a double-antigen sandwich ELISA based on recombinant CCHFV NP derived from strain lbAr10300 ( S segment lineage III ( Africa 3 ) ) . To our knowledge no information is publicly available on the precise nature of the antigen ( native or recombinant antigen , CCHFV strain ) used in the Vector-Best ELISAs . Nevertheless , Vanhomwegen et al . [3] cite a personal communication stating that the manufacturer validated the tests with a serum panel from CCHF cases originating from southwestern Russia . Therefore , we included a CCHFV strain originating from that region ( Drosdov ) in the alignment . Pairwise comparison of the full length NP sequence of the Kosovo Hoti strain with the Afg09-2990 NP ( used as antigen in the BLACKBOX tests ) , the ArD39554 NP ( full virus used in the IIFT tests ) and the Drosdov NP revealed an identity of 96 . 9% , 96 . 9% , and 99 . 4% . When the analysis was restricted to the NP stalk domain , percentages were even higher ( 99 . 2% , 98 . 3% , and 100 . 0% ) . Thus , it is unlikely that the major reason for the observed differences in sensitivity of the BLACKBOX CCHFV IgG ELISA , the VectoCrimean-CHF-IgG ELISA and the CCHFV IgG IIFT is the use of CCHFV antigens resulting from different S segment lineages . Due to the presence of all viral antigens in their native confirmation including potential posttranslational modifications , CCHFV IgG IIFT performed on CCHFV-infected Vero-cells is still the most sensitive method for detecting low levels of CCHFV-specific IgG antibodies in patient sera taken in the early phase of the disease . In addition , the obtained staining pattern allows the internal assessment of signal specificity . Nevertheless , application of this method requires a significant effort in the generation of assay components ( mainly the infection of Vero cells with CCHFV under BSL-4 conditions ) , expensive and sensitive equipment ( i . e . a fluorescence microscope ) and last but not least specialized expertise for test evaluation . In contrast , ELISA components can be produced under BSL-1 conditions and , although assays have to be performed by trained lab personnel , no specialized expertise is necessary to evaluate the result . Beside the competitive sensitivity and specificity of the BLACKBOX CCHFV ELISA tests , there are several advantages of the newly developed tests . First of all , the virus antigen can be produced in large amounts in E . coli; no cost-efficient eukaryotic expression system or cultivation of native virus is necessary . Furthermore , the BLACKBOX CCHFV IgG ELISA requires , due to the co-incubation of recombinant antigen and the patient serum , only one washing step and thus is easy to perform with only short hands-on times . Last but not least , all assay components are routinely produced and quality controlled in a specialized laboratory of the Bernhard Nocht Institute for Tropical Medicine ( Hamburg , Germany ) and thus can be easily made available to the scientific community . In summary , two new serological tests ( BLACKBOX CCHFV IgM and BLACKBOX CCHFV IgG ) for the identification of acute CCHF cases and the performance of seroprevalence studies have been developed . The tests employ recombinant CCHFV NP as antigen and exhibit high reproducibility ( Inter-/inter-assay and inter-laboratory variation ) and competitive assay performance ( sensitivity , specificity ) in comparison with in-house gold standard testing by IIFT and commercially available test kits . | Being endemic in several countries in Asia , Africa , the Middle East and Southeastern Europe , the Crimean-Congo Hemorrhagic Fever Virus ( CCHFV ) is the geographically most widespread tick-borne arbovirus . As evidenced by the recent occurrence of an autochthonous CCHFV infection in Spain , it possesses also a significant potential to spread to as yet non-endemic regions . Due to the severity of the disease caused by this bunyavirus , the lack of specific prophylactic and therapeutic measures and the infection’s epidemic potential , CCHFV was included in the WHO priority list of diseases needing urgent R&D attention , in particular the development and improvement of diagnostic tools . Here we present the development and validation of two novel ELISAs ( BLACKBOX CCHFV IgM , BLACKBOX CCHFV IgG ) for the detection of CCHFV-specific IgM and IgG antibodies employing recombinant CCHFV nucleoprotein ( NP ) as antigen . Test performance in comparison to both in-house gold standard testing ( CCHFV IgM/IgG immunofluorescence test ( IIFT ) ) and commercial ELISA kits ( VectoCrimean-CHF-IgM/IgG; Vector-Best ) was evaluated using a thoroughly characterized serum panel that was obtained from 15 Kosovar patients with an RT-PCR-confirmed CCHFV-infection collected during the years 2013–2016 and that comprised samples from both the acute and convalescent phase of the disease . While both IgM ELISAs , like the CCHFV IgM IIFT , detected CCHFV-specific IgM antibodies in all sera collected during the acute phase of the disease on day 5 after onset of symptoms or later , the BLACKBOX CCHFV IgG ELISA and the CCHFV IgG IIFT were found to be significantly more sensitive than the VectoCrimean-CHF-IgG ELISA in detecting the rising IgG antibody titers in samples collected between days 11 and 19 after onset of symptoms . | [
"Abstract",
"Introduction",
"Materials",
"and",
"methods",
"Results",
"Discussion"
] | [
"medicine",
"and",
"health",
"sciences",
"enzyme-linked",
"immunoassays",
"immune",
"physiology",
"crimean-congo",
"hemorrhagic",
"fever",
"purification",
"techniques",
"immunology",
"tropical",
"diseases",
"parasitic",
"diseases",
"health",
"care",
"nucleoproteins",
"neglected",
"tropical",
"diseases",
"antibodies",
"immunologic",
"techniques",
"protein",
"purification",
"research",
"and",
"analysis",
"methods",
"viral",
"hemorrhagic",
"fevers",
"infectious",
"diseases",
"immune",
"system",
"proteins",
"proteins",
"immunoassays",
"recombinant",
"protein",
"purification",
"biochemistry",
"diagnostic",
"medicine",
"physiology",
"biology",
"and",
"life",
"sciences",
"viral",
"diseases",
"blood",
"donors",
"malaria"
] | 2018 | Sensitive and specific detection of Crimean-Congo Hemorrhagic Fever Virus (CCHFV)—Specific IgM and IgG antibodies in human sera using recombinant CCHFV nucleoprotein as antigen in μ-capture and IgG immune complex (IC) ELISA tests |
The conserved NineTeen protein complex ( NTC ) is an integral subunit of the spliceosome and required for intron removal during pre-mRNA splicing . The complex associates with the spliceosome and participates in the regulation of conformational changes of core spliceosomal components , stabilizing RNA-RNA- as well as RNA-protein interactions . In addition , the NTC is involved in cell cycle checkpoint control , response to DNA damage , as well as formation and export of mRNP-particles . We have identified the Num1 protein as the homologue of SPF27 , one of NTC core components , in the basidiomycetous fungus Ustilago maydis . Num1 is required for polarized growth of the fungal hyphae , and , in line with the described NTC functions , the num1 mutation affects the cell cycle and cell division . The num1 deletion influences splicing in U . maydis on a global scale , as RNA-Seq analysis revealed increased intron retention rates . Surprisingly , we identified in a screen for Num1 interacting proteins not only NTC core components as Prp19 and Cef1 , but several proteins with putative functions during vesicle-mediated transport processes . Among others , Num1 interacts with the motor protein Kin1 in the cytoplasm . Similar phenotypes with respect to filamentous and polar growth , vacuolar morphology , as well as the motility of early endosomes corroborate the genetic interaction between Num1 and Kin1 . Our data implicate a previously unidentified connection between a component of the splicing machinery and cytoplasmic transport processes . As the num1 deletion also affects cytoplasmic mRNA transport , the protein may constitute a novel functional interconnection between the two disparate processes of splicing and trafficking .
The basidiomycete Ustilago maydis is the causal agent of the smut disease on corn ( Zea mays ) . During its life cycle , U . maydis displays yeast-like , haploid cells that divide by budding , and dikaryotic cells that grow as filamentous hyphae . The filamentous stage is initiated by fusion of two yeast-like cells ( sporidia ) and marks the onset of the biotrophic stage in which the fungus depends on its host plant for propagation . The switch to polarized growth to form the elongated filament is indispensable for the successful infection of the host plant . The regulatory circuits that underlie the dimorphic switch and concurrent pathogenic development in U . maydis have been well studied within the past years . Similar to other basidiomycetes , a heterodimeric complex of two homeodomain transcription factors , both encoded by the b-mating type locus , represents the major regulatory instance . The two proteins , termed bE and bW in U . maydis , form a functional heterodimer , but only if they originate from different b-alleles ( e . g . bE1 and bW2 ) . Hyphae formed upon activation of the b-pathway grow unipolar , but only the tip compartment is filled with cytoplasm and contains the two genetically distinct nuclei , while the distal part of the hypha is composed of empty segments separated by evenly spaced retraction septa . Simultaneously with the switch towards filamentous growth , cells become arrested in the G2 phase of the cell cycle . Only after the penetration of the plant surface , this cell cycle block is released , and the “true” filament with multiple septated compartments is developed [for recent review , see 1] . The bE/bW-heterodimer orchestrates a hierarchic , multi-layered transcriptional network . Only few b-responsive genes are direct targets of the bE/bW heterodimer . The majority of the genes is regulated via the b-dependently induced C2H2 zinc-finger transcription factor Rbf1 . Rbf1 is the master regulator for several other transcription factors that coordinate expression of multiple genes associated with cell cycle coordination , morphogenesis and pathogenic development . Deletion of rbf1 prevents the formation of the b-dependent filaments , and ectopic expression of rbf1 is sufficient to induce the dimorphic transition . Thus , rbf1 is both sufficient as well as required for the switch from budding to polarized filamentous growth [2] . Prerequisite for the growth of the filamentous hyphae is the establishment and maintenance of a defined axis of polarity . The filaments expand by polar tip growth , which is dependent on long-distance transport towards the growth cones at the cell apices . This directed transport is facilitated by arrays of polarized microtubules and a highly conserved set of microtubule-dependent kinesin and dynein motor proteins [reviewed in 3] . The cellular cargos that rely on microtubule-based transport include endosomes , peroxisomes and nuclei , but also mRNA , which , as shown recently in U . maydis , is as well instrumental for establishment and maintenance of polarity [4]–[6] . The latter , however , is not a direct microtubule-based cargo , but passively travels on endosomes [7] . In U . maydis , transport is mainly mediated through the concerted action of the plus-end directed UNC104-like Kinesin-3 motor protein Kin3 , which moves endosomes in both directions within the cell along an array of antiparallel microtubules , and the minus-end directed dynein motor protein Dyn1/2 [8]–[11] . The minus-end directed Dyn1/2 is particularly important at the poles of filaments , where unipolar microtubules , with their plus-ends directed to the tip , extend from the antiparallel array [9] , [12] . The conventional kinesin motor protein Kin1 additionally contributes to the establishment of hyphal morphology by transporting Dynein in the direction of the microtubule plus-ends within the hyphal apex where a loading zone for the retrograde transport processes is established [9] , [12] . Other known cargos for the Kin1 motor protein include membranous structures; Kin1 was previously described to be involved in organelle transport [13] and to foster the transport of secretory vesicles into the growing tip [14] . More recently , transport of the fungal-specific class-17 myosin Mcs1 was shown to dependent directly on Kin1 . Mcs1 is attached to vesicles and contains an N-terminal myosin motor domain fused to a chitin synthase region . Anterograde trafficking of Mcs1-positive vesicles requires both microtubules and filamentous actin and depends on Myosin-5 and Kinesin-1 , which cooperate in delivering vesicles to the sites of exocytosis [15]– . In an effort to achieve a better understanding of the complex processes required for the establishment and maintenance of the dikaryotic hyphae in U . maydis , we employed a candidate approach with genes that , based on function or phenotype in other systems , were anticipated to be involved in nuclear migration and determination of cell polarity . Here , we describe the molecular characterization of one of these genes , num1 . The num1 mutation has originally been identified in the basidiomycete Coprinopsis cinerea in a screen for mutants affected in the nuclear migration during the formation of the heterokaryotic hyphae [18] . We show now that Num1 is a homologue of SPF27 , one of the core components of the highly conserved Prp19/CDC5L or NineTeen ( NTC ) splicing complex [19] . Unexpectedly , we identified the conventional Kinesin 1 motor protein Kin1 [13] , [20] to interact with Num1 . Similar phenotypes of Δnum1 and Δkin1 hyphae corroborate a functional interconnection between the two proteins . Our data implicate a previously unidentified connection between a component of the splicing machinery and cytoplasmic ( Kin1-dependent ) transport processes in U . maydis .
In the basidiomycete Coprinopsis cinerea , the num1 gene was identified in a screen for mutants affected in nuclear migration during the initial phase of sexual development [18] . Unlike U . maydis , C . cinerea grows strictly as a filament , and mating is initiated by fusion of two different haploid hyphae harboring compatible alleles of the mating type loci . The dikaryotic filament is then generated upon migration of “donor”-nuclei into the “acceptor”-mycelium . The num1 mutation results in strains that are still able to donate nuclei , but fail to accept nuclei in compatible mating reactions [18] . To examine whether the gene has a conserved function in U . maydis with respect to nuclear migration , we set out to analyze the potential U . maydis homologue . The U . maydis predicted protein Um01682 ( MIPS Ustilago maydis Database ( MUMDB ) , http://mips . helmholtz-muenchen . de/genre/proj/ustilago/ ) exhibits 43% identity and 63% similarity with the C . cinerea Num1 protein , and 33% identity to the human SPF27 protein ( Figure S1 ) . SPF27 homologues are found in all eukaryotic clades , with the exception of the Saccharomycetales ( Figure S2 ) . SPF27 is part of the Prp19/CDC5L complex , which is an integral component of active spliceosomes and required for intron removal [21] . Similar to the SPF27 homologues , the U . maydis protein harbors a BCAS2 domain ( breast carcinoma amplified sequence 2 ) with a so far unknown function and a classical , basically charged nuclear localization signal ( Figure S1 ) [22] . To determine the function of Num1 in U . maydis , the gene was deleted in the strains FB1 , FB2 , SG200 and AB31 . FB1 ( a1 b1 ) and FB2 ( a2 b2 ) are compatible wild-type strains that , upon mating , form filaments on solid media supplemented with charcoal . SG200 ( a1 mfa2 bE1 bW2 ) forms such filaments without a mating partner , as the strain harbors a compatible combination of the bE1 and bW2 genes [23] . For both FB1/FB2 and SG200 , cell division upon filament formation is stalled in axenic culture , and hyphal proliferation depends on the plant host . Polarized growth in axenic culture can be best monitored in strain AB31 ( a1 Pcrg1:bE1/Pcrg1:bW2 ) , which harbors a set of compatible bE1 and bW2 genes under control of the arabinose-responsive Pcrg1-promoter [24] . Upon arabinose-induced expression of bE1/bW2 , AB31 switches from budding- to polarized growth and forms long filaments reminiscent of those formed after fusion of compatible FB1 and FB2 sporidia [2] , [24] . Yeast-like cells of AB31Δnum1- and SG200Δnum1 showed no obvious phenotype when grown in axenic culture in complete or minimal media ( Figure S3 ) . In plant infection assays , the mixture of the compatible strains FB1Δnum1 and FB2Δnum1 as well as SG200Δnum1 showed reduced disease symptoms compared to the respective wild-type strains; both the number of plants with tumors as well as the size of tumors decreased ( Figure S3 ) . However , microscopic observation of hyphae after plant infection did not reveal major morphological differences to wild-type strains ( Figure S3 ) . The cell-cell fusion of compatible sporidia to generate the dikaryotic hyphae , which can be visualized by the appearance of “fuzzy” colonies on charcoal containing media , is not altered by the num1 mutation ( Figure S3 ) . However , the phenotype of conjugation hyphae upon treatment with synthetic pheromone as well as the phenotype of the dikaryotic hyphae showed considerable alterations from wild-type strains . Pheromone treatment of Δnum1 strains led to the formation of branched conjugation tubes that are never observed in wild-type strains . Dikaryotic hyphae of Δnum1 strains appeared thicker , branched , and fused sporidia displayed bipolar growth; in wild-type strains , however , the fusion event is initiated only from a single cell pole ( Figures S4 and S5 ) . To analyze the phenotype of the hyphae in more detail , we employed strain AB31 and induced the formation of the filament via the expression of bE1/bW2 . Similar to the dikaryotic hyphae , AB31Δnum1 filaments displayed severe alterations . 95% of the filaments were significantly shorter than wild-type AB31 filaments ( average length of 68 . 9 µm compared to 121 . 7 µm , t-test: p = 6 . 35×10−46 ) . Filament formation upon b-induction in AB31 is initiated at one of the cell poles of the sporidium; the resulting hyphae grow unbranched and strictly unipolar ( Figure 1A ) . In AB31Δnum1 , hyphae generally exhibited an irregular and curved morphology ( Figure 1B , Table 1 ) . In 18% of the sporidia , hyphae initiated at both cell poles ( bipolar , Figure 1C , Table 1 ) , and 4 . 4% exhibited a branched morphology ( N = 461 ) ( Figure 1D , Table 1 ) . In addition , 21 . 2% of the AB31Δnum1 filaments had delocalized septa , and 2 . 4% of the hyphae generated empty compartments at the cell tips ( Figure 1E and F , Table 1 ) . Septa in AB31Δnum1 could be visualized by different chitin-stains , as Calcofluor [25] , Congo Red [26] or fluorescein-conjugated wheat germ agglutinin ( WGA ) [27] ( Figures 1E and S6; see also Figure S6 legend ) . However , in contrast to septa in the dikaryotic hyphae , septa in AB31Δnum1 were not always visible in the DIC channel , and staining with fluorescein-WGA revealed extensive chitin accumulations or ring-like chitin structures throughout the hyphae , indicating that these septa or septa-like structures may be structurally different from wild-type septa . To ensure that the observed phenotype is due to the num1 deletion , a construct harboring a num1:eGFP fusion gene under control of the arabinose inducible Pcrg1-promoter was introduced in single copy into the cbx locus . Strain UNK220 ( AB31Δnum1 Pcrg1:num1:egfp ) was grown in arabinose-containing media to induce both the bE1/bW2 heterodimer as well as the num1:eGFP gene; the induced filaments were indistinguishable from that of the wild-type AB31 strain ( Figure S7 ) . Num1 has striking similarities to the human BCAS2/SPF27 protein that has been identified as one of the four core components ( Prp19 , PLR1 , CDC5L and SPF27 ) of the Prp19/CDC5L spliceosome-associated complex , commonly termed NineTeen Complex ( NTC ) ( Figure 2A ) . To assess whether Num1 is a structural component of the putative NTC complex in U . maydis , we employed a directed yeast two-hybrid approach using the full-length Num1 protein as bait as well as full-length and C- and N-terminal versions of the Prp19 ( Um10027 ) and CDC5L-homologues ( termed Cef1 , Um04411 ) as prey , respectively . Consistent with the previously identified interactions in S . pombe [28] , Num1 interacts with the Prp19 N-terminus as well as the Cef1 C-terminus ( Figure 2B ) . The interaction of Num1 with Prp19 and Cef1 was verified by in vivo co-immunoprecipitation . num1 was replaced by an eGFP-tagged version of the gene in strain AB31 ( a2 , Pcrg:bW2 , bE1 ) , and additionally prp19 or cef1 were exchanged with HA-tagged versions , resulting in strains UNK200 ( AB31 num1:3×eGFP:hygR , prp19:3×HA:cbxR ) and UMO8 ( AB31 num1:3×eGFP:hygR , cef1:3×HA:cbxR ) , respectively . All genes were introduced via homologous recombination into their native loci to ensure native protein levels . All fusion proteins were shown to be functional ( Supporting Text S1 , Figure S8 ) . By use of an anti-HA-antibody , Num1:3eGFP was specifically co-precipitated with both Prp19:HA and Cef1:HA in protein extracts from strains UNK200 and UMO8 ( Figure 2C ) . The interaction of Num1 with Prp19 and Cef1 was further corroborated by their localization . Num1:3eGFP and Prp19:RFP or Cef1:RFP were co-expressed in strain AB31 , and all proteins co-localized in the nuclei ( Figure 2E and F ) . While for Prp19:RFP and Cef1:RFP the fluorescence signal was exclusively localized in the nucleus , Num1:3eGFP was also detected in the cytoplasm ( Figure 2D ) . Transformation of AB31Δnum1 with a num1:eGFP derivative harboring a mutated nuclear localization sequence ( KKRK to AAAA ) did not lead to complementation of the Δnum1 phenotype , and the fusion protein localized predominantly in the cytoplasm ( Figure S7 ) , which implies that nuclear localization of Num1 is required for function of the protein . In addition to their function in mRNA splicing , NTC and NTC-associated proteins are known to be involved in several other cellular processes . In S . pombe and S . cerevisiae , Cdc5 and Cef1p , respectively , were originally identified as cell division cycle mutants that are required for G2 progression and mitotic entry [29]–[31] , and Prp19p has been isolated in S . cerevisiae in a screen for mutants conferring sensitivity to DNA damaging agents [32] . It has been proposed that Prp19p homologues play a direct role in DNA damage repair that is conserved from yeast to human [33]–[38] . To define whether Num1 acts in line with previously known functions of the NTC-complex , we investigated cell cycle regulation and the response to DNA damage in AB31Δnum1 . In AB31 , the filaments generated after induction of the bE1/bW2 heterodimer are arrested in the G2 phase of the cell cycle , harboring a single haploid nucleus that is positioned in the tip compartment [39] . In filaments of AB31Δnum1 , however , deviations from this pattern became evident . 18% of the filaments ( N = 100 ) exhibited two ( or more ) nuclei within a single hyphal compartment; additionally , compartments devoid of any nuclei were generated , most likely due to the aberrant formation of septa ( Figure 3A ) . In accordance with this observation , FACS analysis of AB31 cells revealed that upon prolonged induction of the bE1/bW2-heterodimer cells with a 2C DNA content accumulated , whereas in AB31Δnum1 , an increased number of cells with both 1C and 2C DNA content was observed ( Figure S9 ) , indicative for a deregulated cell cycle . Taken together , our results implicate that the num1 deletion affects both cell cycle regulation as well as cell division . Consistent with the function of the NTC in response to DNA damage , we found that Δnum1 cells exhibited increased sensitivity to UV irradiation and DNA-damaging agents . The survival rate upon UV-treatment was reduced approximately by a factor of 10 in AB31Δnum1 ( Figure 3B ) . Similarly , treatment with phleomycin , causing double-strand breaks in DNA [40] or hydroxyurea , which triggers stalling of the replication fork during DNA synthesis [41] , led to reduced survival of Δnum1 cells ( Figure 3B ) . To assess whether the Δnum1 mutation affects splicing , we employed next-generation RNA sequencing ( RNA-Seq ) to determine the transcriptome of strains AB31 and AB31Δnum1 . bE1/bW2 expression was induced for eight hours; at this time point , in both strains a comparable quantity of filamentous cells was observed . The expression levels of bW , of the central regulator rbf1 ( a direct b-target gene ) and of num1 were monitored by quantitative RT-PCR analysis to ensure similar expression levels ( Figure S10 ) . For both AB31 and AB31Δnum1 , RNA-Seq experiments from three independent biological samples were performed , and sequences were compared to the manually curated U . maydis gene models of the MIPS Ustilago maydis Database ( MUMDB ) . Splicing efficiency was calculated from the normalized read number within introns in relation to the normalized read number of exon regions of a gene ( for details , see Materials and Methods ) . Based on 2142 intron sequences , introns of AB31 genes were spliced with an average efficiency of 0 . 93 ( ±0 . 13 ) ; splicing efficiency in AB31Δnum1 was significantly lower ( factor −1 . 44 , t-test p = 6 . 22*10−111 ) , with an average of 0 . 80 ( ±0 . 23 ) . The splicing defect observed by RNA-Seq was verified for a small set of genes by qRT-PCR ( Figure S11 ) . In general , we noticed a wide variation in splicing efficiency of different introns in response to the num1-deletion . For 80 . 2% of the introns ( 1717 of 2142 ) , the intron retention rate ( fraction of RNA in intronic sequences in relation to exonic sequences ) in AB31Δnum1 was at least twice as high as in the wild-type strain , indicating that Num1 is required for splicing in general . We noticed , however , that for few introns ( 83 , allowing a 10% threshold ) splicing was not affected , while other introns were barely spliced at all in AB31Δnum1 . Differences in intron retention rates were specific for discrete introns , and not for particular transcripts , as we observed that in genes with more than one intron separate introns were spliced to different extent ( Table S1 , examples are shown in Figure S12 ) . It has been proposed that the function of the NTC with regards to cell cycle regulation or DNA damage repair may result from the decreased splicing efficiency of intron-containing genes involved in recombination , repair , cell cycle or chromosome segregation [42] . We did not observe an enrichment of genes with high intron retention rates in any “category for functional annotation” ( FunCat ) . However , inspection of genes in the FunCats “cell cycle” and “DNA repair” ( Table S2 ) revealed that for individual genes involved in DNA repair splicing is impaired , as , for example , for rad51 ( um03290 ) , for the gene for the repair exonuclease Rec1 ( um11016 ) , or for mms2 ( um10097 ) , involved in the error-free post-replication repair . Similarly , splicing of genes involved in cell cycle control , as cks1 ( um03210 ) , encoding a regulatory subunit of cyclin-dependent kinases , or the gene for the cyclin dependent kinase 1 ( cdk1 , um 10705 ) is affected in AB31Δnum1 ( Table S2 , Figure S13 ) . We have shown before that cell cycle and hyphal development are controlled by a cascade of transcription factors [2] . Intriguingly , although most U . maydis genes contain no introns , both bW and bE carry single introns; rbf1 harbors even four introns . It could be well possible that the observed phenotype of the Δnum1 strains is caused by inefficiently spliced genes of the master regulators for filamentous growth and pathogenic development . In AB31 , the bW2-allele is expressed without the intron sequence [24] , but we observed that bE1 is indeed spliced less efficiently in AB31Δnum1 ( Table S3 ) . However , all direct b-target genes were induced to a similar extent in AB31Δnum1 compared to AB31 ( Table S4 ) , which implies that expression of the bE and bW genes in the Δnum1 background is sufficient to provide full functionality of the bE/bW heterodimer . In accordance with the unaltered functionality of bE/bW , the expression of the direct target gene rbf1 , was not altered in AB31Δnum1 ( Figure 4C ) ; however , all four introns were spliced less efficiently . Under the assumption that splicing of the introns occurs independently from each other , only in 45% of the mature mRNA in AB31Δnum1 all introns were spliced correctly , compared to 80% in AB31 wild-type cells ( Figure 4A , Table S3 ) . The splicing defect in rbf1 was validated via RT-PCR and qRT-PCR analyses ( Figure S14 ) , showing additional unspliced variants and increased intron retention , respectively , in AB31Δnum1 . Consistently , we observed that the abundance of the Rbf1 protein in AB31Δnum1 is reduced to about 30% of wild-type level ( Figure 4B ) , although qRT-PCR analysis showed that the total level of the rbf1-transcript was not altered ( Figure 4C ) . Rbf1 controls the expression of biz1 , hdp1 and hdp2 , encoding transcription factors that are collectively involved in cell cycle regulation , filamentous growth and pathogenic development [43 , Scherer , Vranes , Pothiratana and Kämper , unpublished] . In accordance with the reduced Rbf1 levels , RNA-Seq analysis revealed that biz1 as well as hdp2 expression was significantly down-regulated in AB31Δnum1 , which was verified by qRT-PCR ( Figures 4D and S15 , Table S4 ) . hdp1 splicing efficiency was reduced in AB31Δnum1 from 0 . 83 to 0 . 14 ( Table S3 ) ; interestingly , hdp1 RNA-levels were increased about 2-fold , which may argue for a self-regulation of the gene ( Figures 4D and S15 , Table S5 ) . In total , we observed an at least two-fold down-regulation for 228 genes in AB31Δnum1; intriguingly , 89 of these genes overlap with a set of 351 genes that has been identified as down-regulated in response to a deletion of rbf1 in AB31 ( Figure 4E , Table S5 , [2] ) . As rbf1 is one of the major regulators implicated in filamentous growth and cell cycle regulation , it was tempting to speculate that the reduced splicing efficiency of the gene may be the cause for the observed phenotype in Δnum1-mutants . To test this hypothesis , we exchanged the rbf1 gene with an intron-free derivative in strains AB31 and AB31Δnum1 . However , we observed that the intron-free copy of rbf1 is not capable to substitute for the native gene; despite wild-type-like expression levels the strains did not grow filamentous , nor were the known rbf1-target genes induced ( Figure S16 ) . Obviously , introns are essential for the expression of the Rbf1-protein . As a consequence , expression of the rbf1 cDNA did not rescue the Δnum1 phenotype either . In summary , we conclude that Num1 is a structural component of the conserved Prp19/CDC5L complex and fulfills common functions that have previously been associated with this complex , as cell cycle control , cellular response to DNA damage , and splicing . Our data indicate that the observed phenotype in U . maydis hyphae may be at least in part the result of the splicing defect of genes involved in cell cycle control , DNA damage repair and of central components of the regulatory cascade controlling filamentous growth and pathogenic development . To further elaborate the molecular function of the Num1 protein , we employed the yeast two-hybrid system . Proteins identified as potential interactors for Num1 are shown in Table S6 . Consistent with a putative function of Num1 within the Prp19/CDC5L complex , the CDC5L homologue Cef1 was isolated as an interacting protein . Unexpectedly , we identified various interactions with proteins involved in cellular transport processes . Um10158 is an adaptin-like protein from clathrin/coatomer-adapter complexes ( IPR002553 ) mediating endocytic protein transport between ER and Golgi . Um03539 and Um11510 contain a BAR-domain ( IPR004148 ) and a BRO1-domain ( IPR004328 ) , respectively , and are thus likely involved in intracellular vesicular transport processes and protein targeting to endosomes or the vacuole . The conventional kinesin motor protein Kin1 [formerly named Kin2 , 20] functions in long distance trafficking within fungal hyphae as it transports Dynein along the microtubule cytoskeleton towards their plus-ends directed to the hyphal tips [9] , [14] . We were not able to generate deletion strains for um10158 , indicating that the respective protein might be of essential function . Deletion analysis for um03539 and um11510 in strain AB31 revealed no obvious phenotype with respect to polar growth and vesicular movement ( data not shown ) , indicating that the respective proteins have no function with respect to Num1-associated processes . For kin1 , however , it had already been demonstrated that the gene has impact on hyphal morphology [14 , see below] . To confirm the interaction of Num1 with Kin1 , co-immunoprecipitation analyses were carried out . A full-length Num1:Myc-tagged protein and a Kin1:HA-tagged protein fragment encompassing amino acids 650–968 were generated by coupled in vitro transcription/translation , and Num1:Myc was specifically co-precipitated from mixtures of Kin1:HA with an anti-HA antibody ( Figure 5A ) . By use of the yeast two-hybrid system we mapped the interaction domain of Num1 ( encompassing amino acids 67 to 147 ) with the C-terminal Kin1 domain used in the co-immunoprecipitation , which provides additional evidence for a specific interaction with the Kin1 motor protein ( Figure S17 ) . The interaction was further verified in vivo , using U . maydis strain UNK197 that harbors num1:3eGFP and kin1:HA fusion genes integrated into their native loci by homologous recombination , respectively . By use of dithiobis[succinimidyl]-propionate ( DSP ) to crosslink proteins in extracts prior to immunoprecipitation , the Num1:3eGFP protein was efficiently co-precipitated with the Kin1:HA protein using anti-HA antibodies ( Figure 5B ) . Without cross-linking of the proteins , only a faint signal was detectable ( not shown ) , suggesting that the interaction is only transient . Taken together , we conclude from our data that Num1 interacts with the motor protein Kin1 . In addition to the physical interaction , several lines of evidence argue for a functional connection between Num1 and Kin1 . Induced expression of bE1/bW2 in a Δkin1 strain leads to short curved hyphae with delocalized septa and the formation of bipolar filaments [14] . Similarly , the induction of hyphal growth in AB31Δnum1 leads to irregular or bipolar filaments with an aberrant insertion of septa ( Figure 1 , Table 1 ) . In strains deleted for kin1 the number of vacuoles is increased , while size is reduced [13] . We observed a similar phenotype in strains SG200Δnum1 ( Figure 6A ) . To analyze the functional relation between the two proteins further , we generated strains deleted for both genes . Filaments induced in AB31Δnum1/Δkin1 displayed a more severe phenotype with respect to polarized growth , as hyphae exhibited a very irregular , swollen morphology , and also the number of branched hyphae and hyphae with an altered septation pattern was significantly higher than in AB31Δnum1 or AB31Δkin1 ( Figure S18 ) . Infection experiments with SG200Δnum1 , SG200Δkin1 and SG200Δnum1/Δkin1 deletion strains revealed that the virulence of the double deletion strain is significantly reduced when compared to the respective single deletion strains ( Figure S18 ) . The more severe phenotype excludes that both genes are in pure epistatic interaction; besides their function within a common complex , both proteins must fulfill functions that are independent from each other . The deletion of kin1 results in a loss of rapid bi-directional trafficking of early endosomes . Early endosomes visualized by a GFP-fusion with the endosomal t-SNARE Yup1 ( Yup1:eGFP ) [4] accumulate at the hyphal tips of kin1-deletion mutants [9] , while in the wild-type situation Yup1:eGFP labeled vesicles shuttle bi-directionally throughout the hyphae along microtubule tracks [4] . The anterograde transport of early endosomes is mediated via the motor protein Kin3 ( directed to the microtubule plus-ends at the hyphal tip ) , whereas the retrograde transport is accomplished by Dynein1/2 [8] , [10] , . Since the Kin1 motor protein transports Dynein toward the hyphal tips , the accumulation of early endosomes at the tip of Δkin1 hyphae can be explained indirectly by the failure of the Kin1-dependent anterograde transport of Dynein [9] . The deletion of num1 leads to a comparable scenario in the distribution of early endosomes , which appeared to cluster at hyphal tips and the distal pole of the hyphae as well as around the delocalized septa frequently found in Δnum1-hyphae ( Figure 6B ) . Endosomal motility was monitored in strain AB33 , in which the b-genes are expressed under control of the nitrogen-responsive Pnar1-promoter [44] . Filamentous growth in AB33 is repressed by ammonium and can be elicited in nitrate-containing media . In AB33Δnum1 filaments , long-distance movement of early endosomes was constrained , and velocity of the vesicles was reduced to a speed of 1 . 14 µm/sec in AB33Δnum1 , in contrast to 1 . 46 µm/sec in wild-type hyphae ( t-test p = 0 . 005 ) . In AB33Δnum1 filaments with aberrant septa , long-distance trafficking of endosomes was entirely abolished , and only residual movement of vesicles was observed ( Table 2 , Videos S1 and S2 ) . Next we addressed the question whether the altered movement and positioning of early endosomes in Δnum1 strains may be caused by a distortion of cytoskeletal elements . Neither the arrangement or abundance of microtubules ( visualized by an α-tubulin mCherry fusion , Tub1:mCherry ) nor of actin patches ( visualized by LifeAct:YFP [45] ) was altered in the Δnum1 mutant strain ( Figure S19 ) . To analyze the orientation of the microtubule cytoskeleton , we used strains expressing an α-tubulin GFP fusion [14] together with the microtubule plus-end marker Peb1 , fused to RFP , which localizes to growing microtubule plus-ends [46] . As previously described [46] , in wild-type interphase sporidia the majority of microtubule plus-ends is oriented towards the budding cell and the opposing cell pole of the mother cell; in Δnum1 sporidia , the orientation of the microtubule cytoskeleton was not considerably altered ( Figure S20 A , B ) . Similarly , in hyphae the majority of microtubule plus-ends is oriented towards the cell poles , i . e . hyphal tips and basal retraction septa [9] . In Δnum1-mutants , we observed microtubule plus-ends also in close vicinity to the delocalized septa ( Figure S20 C–E ) . Apparently , endosomes accumulate at microtubule plus-ends in num1-deletion mutants , supporting the Kin1-dependent effect on endosomal transport . In some hyphae of the num1-deletion strain we observed the appearance of mitotic spindles , which supports the defect in cell cycle regulation . To determine whether the effect on endosomal trafficking in AB33Δnum1 strains is attributed to an aberrant anterograde transport of Dynein towards the hyphal apex , as it was suggested for Δkin1 mutants , we visualized the subcellular distribution of Dynein using an N-terminal 3eGFP:Dyn2 fusion protein [9] . In wild-type filaments , 3eGFP:Dyn2 fusion proteins form comet-like motile structures that accumulate at microtubule plus-ends in hyphal tips , where they reverse direction and move into retrograde direction [9] . In AB33Δnum1 filaments , however , the number of cells harboring an apical accumulation of 3eGFP:Dyn2 was significantly reduced ( t-test p = 0 . 002 ) ; concomitantly , in most cells the 3eGFP:Dyn2 signal in the cytoplasm was higher than in wild-type filaments ( Figure 6C ) . As Kin1 affects motility of early endosomes only indirectly ( via the anterograde transport of Dynein ) , we set out to analyze the movement of a direct Kin1-cargo . The myosin chitin synthase Mcs1 [15] localizes to Kin1-dependently transported vesicles [17] . It has been described before that 3eGFP:Mcs1 fusion proteins localize to the poles of growing sporidia , where it is secreted and participates in the synthesis of chitin [15] . Δnum1-mutants showed reduced Mcs1-accumulations at the growth region in budding cells , which coincides with the distribution of Mcs1 in kin1-deletion mutants [17] ( Figure S21 A , B ) . In AB33 filaments , 3eGFP:Mcs1 localizes to distinct foci close to the cell membrane and forms a gradient towards the growth zone within the hyphal apex or localizes to the basal retraction septum , i . e . to sites of chitin synthesis . In contrast , in AB33Δnum1 hyphae no tip-directed gradient was obvious and less foci were observed at the cell membrane ( Figure S21 C , D ) , indicating an impaired delivery of Mcs1-containing vesicles . In many cases , Mcs1 was not observed at delocalized septa in Δnum1-mutant hyphae . In summary , U . maydis strains deleted for either num1 or kin1 show similar phenotypes with respect to hyphal morphology and septation , formation of vacuoles , endosomal trafficking as well as the localization of the Kin1-cargos Dynein and Mcs1 , which strongly corroborates a functional interrelation of Num1 and Kin1 . Recently , it has been shown that the RNA-binding protein Rrm4 is associated with early endosomes [7] . Rrm4 is required for cytoplasmic transport of mRNAs in U . maydis filaments [47] . We deleted num1 in an AB33 strain expressing an Rrm4:eGFP fusion protein . In the wild-type background vesicles associated with eGFP-labeled Rrm4 shuttled within the hyphae to the distal and proximal ends . In contrast , movement in Δnum1 was reduced , and similar to Yup1-labeled early endosomes , Rrm4-containing particles frequently accumulated in clusters at hyphal tips or around delocalized septa ( Figure S22 , Videos S3 , S4 ) , indicating that num1 is required for cytoplasmic mRNA transport . To address the function of Num1 in filamentous fungi in general , we deleted a gene encoding a protein with high similarities to Num1 ( AN4244 , 49% similarity , 30% identity ) in the ascomycete Aspergillus nidulans ( strain TN02A3 ) . The deletion mutants displayed smaller colonies with a reduced rate of conidiospore formation . Germination of spores was drastically reduced in the ΔAN4244 mutant strains: after eight hours of incubation at 37°C , only 5% of the spores germinated , compared to 98% in the wild-type strain . Similar to the phenotype observed in U . maydis , hyphae from germinating ΔAN4244 conidia were shorter and thicker than those of the wild-type strain . Staining with CellTracker Blue revealed that in the mutant the number of vacuoles was increased , while size was reduced ( Figure 7A ) . In contrast to U . maydis , A . nidulans hyphae have cell compartments with multiple nuclei . In wild-type hyphae , these nuclei are evenly distributed . In ΔAN4244 , however , nuclei were often found in clusters ( Figure 7B ) . A similar phenotype has been observed earlier for strains deleted for KinA , the homologue of the U . maydis Kin1 [48] . To test a potential functional relation between AN4244 and KinA in A . nidulans , we analyzed , in analogy to U . maydis , the distribution of Dynein ( NudA in A . nidulans ) . We deleted AN4244 in strain LZ12 , which harbors a GFP:NudA fusion protein . Similar as described for the U . maydis hyphae , GFP:NudA was predominantly localized at the tip of hyphae in strain LZ12 , while the apical accumulation was reduced in LZ12ΔAN4244 ( Figure 7C ) . Taken together , our data suggest that the A . nidulans AN4244 has a conserved function as the U . maydis protein with respect to polarized growth and intracellular transport processes .
Num1 shows significant similarities to SPF27 , and its interaction with Prp19 and Cef1 is in accordance with the interactions within the NTC complex demonstrated in S . pombe [28] , humans [21] and plants [49] . The NTC was initially identified in S . cerevisiae as a complex containing Prp19 and at least seven other components [50] , most of which are conserved from yeast to human . The NTC is a non-snRNP ( small nuclear ribonucleoprotein particle ) multi-protein complex , and as a subunit of the spliceosome it is required for intron removal during pre-mRNA splicing . The complex associates with the spliceosome during or after dissociation of U4 snRNP and participates in the regulation of conformational changes of core spliceosomal components , stabilizing RNA-RNA- as well as RNA-protein interactions [reviewed in 51] . The only known enzymatic function of the NTC is a conserved U-box domain with E3-ubiquitin ligase activity of the Prp19-protein [52] , [53] . It was conclusively shown that individual core components of the NTC , as Prp19 , Cef1/Cdc5 in S . cerevisiae and S . pombe , or the S . pombe SPF27 homologue Cwf7 independently affect splicing [42] , [50] , [54]–[59] . However , it is mostly unresolved to which extend single components of the NTC influence splicing of individual introns . We now show that one of the core components of the NTC , the SPF27 homologue Num1 , has a global effect on splicing in U . maydis , as retention rates of nearly 80% of all introns were significantly increased in num1-deletion mutants . We observed variations in intron retention between different transcripts in response to the Δnum1 mutation . Similar observations have been made in S . cerevisiae on a genome-wide scale for depletion of individual components of the spliceosome , including the NTC-components Prp19 and Prp17 . The resulting splicing defects displayed considerable differences in different mRNAs , indicating that individual splicing factors affect a discrete set of genes [60] . It is well possible that structural alterations of the spliceosome complex after deletion of num1 result in altered recognition/affinity of specific introns , probably caused by their different GC content or secondary structures . Following the conclusions made in S . cerevisiae , we favor the possibility that deletion of num1 leads to an altered integrity of the NTC , resulting in a reduced activity and/or specificity and hence a diminished splicing efficiency . However , we cannot rule out that Num1 may recruit splicing enhancer- or inhibitor-proteins that might affect splicing of only a subset of introns . In addition to disturbed polarity we frequently observed delocalized septa in num1-deletion strains; in addition , chitin-staining indicate structural differences to wild-type septa . Formation of septa requires a highly coordinated and dynamic positioning of septins into ring-like structures [61] , a process tightly regulated by the small GTPase Cdc42 , its guanine exchange factor Don1 and the protein kinase Don3 [62]–[64] as well as by formins or Cdc4 , an essential light chain of the type II myosin motor protein [65] . The genes encoding the septins Cdc3 ( um10503 ) and Sep3 ( um03449 ) , but also Cdc4 ( um11848 ) display significantly increased intron retention rates ( Table S1 ) and the expression of don3 ( um05543 ) is reduced more than 2-fold in the num1-deletion strain ( Table S4 ) , which altogether might rationalize the formation of delocalized septa . We found that Num1 affects splicing of several genes of the functional class “cell cycle” , including genes for the catalytic and regulatory subunits of cyclin-dependent kinases , for mitotic cyclins as well as for γ-Tubulin ( Table S2 ) . Similarly , several genes grouped in the functional class “DNA-repair” ( e . g . for the DNA repair recombinase Rad51 , the nucleotide excision machinery component Rad1 and the repair exonuclease Rec1 ) show increased intron retention rates ( Table S2 ) . The synergistic effect of several genes with reduced splicing efficiency ( and the resulting reduced activity ) within the same functional category may be sufficient to explain the effects on the corresponding cellular processes . However , various components of the NTC have been found to be directly involved in DNA damage repair . Initially , the resistance to the interstrand cross-linking reagent psoralen in a S . cerevisiae Prp19 mutant strain [32] has been attributed to the splicing defect of the intron containing RAD14 gene [66] . Subsequently , it has been conclusively shown that the DNA damage phenotype is independent from the splicing defect [42] . In addition , for Prp19 several interaction partners have been identified which belong into the functional classes “cell cycle” , “chromatin structure” and “DNA repair” rather than pre-mRNA processing . For example , the human Prp19/Pso4 directly interacts with terminal deoxynucleotidyl-transferase ( TdT ) , which is involved in DSB repair [67] . The human Cef1-homologue CDC5L interacts with the Werner syndrome protein WRN , a DNA helicase that functions during homologous recombination in response to DNA-damage [35] and also with the non-homologous end-joining ( NHEJ ) factor DNA PKcs [68] . Thus , it is well possible that also in U . maydis the increased UV-sensitivity of Δnum1 mutants results from a direct function of Num1 in a DNA damage repair pathway . Similarly , disturbed cell cycle and cell division can be explained as indirect effect of a disturbed DNA-damage repair . Interaction of the human CDC5L with the cell cycle checkpoint kinase ATR ( “ataxia-telangienctasia and Rad3-related” ) was shown to be required for activation of the ATR kinase as well as for further downstream effectors and mediators of ATR function as the checkpoint kinase CHK1 [69] . In U . maydis , the homologues of ATR1 and CHK1 are known to be critical cell cycle regulators in addition to their decisive roles in the DNA-damage response [70] , [71] . We have shown previously that cell cycle and polarized growth in U . maydis is regulated by a cascade of transcription factors controlled by the bE/bW-heterodimer [2] . Several of the transcription factors of this b-cascade were found to have increased intron retention rates or a differential expression pattern in AB31Δnum1 and might collectively contribute to the effects on cell cycle control and polar hyphal growth . For example , splicing efficiency of the hdp1 gene , which encodes a homeodomain transcription factor required for filamentous growth and cell cycle control ( Pothiratana and Kämper , unpublished data ) , was reduced from 83% in wild-type to 14% in the num1-deletion strain . More importantly , all four introns of rbf1 are spliced less efficiently in the num1-deletion strain . As a consequence , protein levels of Rbf1 , the master regulator for hyphal development , are significantly decreased in Δnum1 strains . In accordance with the reduced Rbf1 protein levels in Δnum1 strains , we observed the down-regulation of rbf1-dependently expressed genes . Among these are the rbf1-dependently expressed genes encoding the transcription factors Biz1 and Hdp2 , which were previously described as cell cycle regulators and factors for filamentous growth and pathogenic development [43; Scherer and Kämper , unpublished data] . Conclusively , we can assume that the reduced splicing efficiency of the rbf1 gene and in turn the deregulation of rbf1-dependently expressed genes contributes to hyphal morphology and the deregulated cell cycle in num1-deletion mutants . A direct proof of this hypothesis by replacing the rbf1 open reading frame with a cDNA copy failed . The intron-free copy , although expressed to the same level as the intron-containing copy , was not able to provide wild-type Rbf1 function , even when introduced into wild-type cells . Such a reduced expression of cDNA copies in relation to intron-containing genes has been described frequently before , and in various cases this has been related to direct effects of introns to translational efficiency [reviewed in 72] , [73] . Aberrant or incomplete splicing events of the genes encoding motor proteins by themselves would be an explanation for the impaired endosomal trafficking processes observed in num1-deletion strains . However , none of the genes for the major transport-mediating motor proteins ( kin1 , kin3 , dyn1/2 ) contains introns or show altered expression levels in the Δnum1-mutant . Although we cannot rule out an effect caused by aberrant splicing of genes for unknown components of the transport machinery , we favor a specific role of the Num1 protein in intracellular trafficking processes , based on the observation that ( a ) Num1 is localized in distinct foci in the cytoplasm , implicating functions beyond its role as a splicing factor , and ( b ) Num1 interacts with the Kin1 motor protein , which is involved in long-distant transport processes during filamentous growth [9] , [12] , [14] . The physical interaction of Kin1 and Num1 is in line with the similar phenotypes of the two mutant strains with respect to cell morphology , determination of polarity and septation , and morphology of vacuoles . Δnum1 and likewise Δkin1 hyphae display impaired trafficking of early endosomes and reduced apical localization of Dynein , two interconnected phenotypes , as the transport of early endosomes relies on Dynein and the anterior transport of Dynein depends on Kinesin 1 . In addition , the distribution of the direct Kin1 cargo Mcs1 is altered in both Δkin1 and Δnum1 mutants , which argues for a direct connection of Kin1 and Num1 . However , we can rule out that Num1 is required for all Kin1-related processes , as ( a ) both mutations appear not to be epistatic , and ( b ) the num1 and kin1 mutant strains have overlapping , but not identical phenotypes; for example , haploid Δkin1 strains are defective in mating [20] , while Δnum1 strains are not . This is anticipated , as each of the proteins participates in different cellular processes . The conventional kinesin is best known for its function in cellular transport processes . However , conventional kinesin has also been implicated in cell signaling , or to serve as bona fide , static anchor for mRNA–protein complexes or vesicular compartments , or to participate in the phosphorylation of cargo proteins to which it binds [reviewed in 74] . As outlined above , also SPF27 ( or the NTC ) participates in a wide spectrum of processes that , however , show no overlap with Kin1-dependent processes . Interaction of SPF27 homologues with conventional kinesin motor proteins is not specific to U . maydis , but appears to be conserved between basidio- and ascomycetes , as in Aspergillus nidulans the phenotype of deletion mutants of the SPF27 homologue AN4244 or the Kin1 homologue KinA [48] are comparable to those observed for U . maydis . The association of a conventional kinesin motor protein with a component of the spliceosome-associated NTC has not been described so far and implies a novel function for SPF27-homologous proteins . The question is whether the Num1 protein fulfills dual functions , ( a ) during splicing as an NTC-component in the nucleus and ( b ) during Kin1-dependent intracellular trafficking processes in the cytoplasm , or whether these two disparate mechanisms are connected by one single function . Discrete cytoplasmic functions may serve to regulate the activity of Kin1 , as described in neuronal development for the interaction of Kinesin-1 and JIP3 , an interactor of the c-Jun N-terminal kinase [75] . Alternatively , Num1 could function as a scaffold or adapter molecule , providing stability to the motor protein/cargo-complex and/or to the NTC . In both scenarios , deletion of num1 could alter the specificity/activity of Kin1-mediated transport processes , resulting in reduced motility of early endosomes and by that influence polar growth . We have not been able to visualize Num1 proteins moving or co-migrating with Kin1 along microtubule tracks . As GFP-Kin1 shows a strong cytoplasmic background , visualization of Kin1 molecules moving on microtubules has not been achieved yet [14] , [76] . Taken into account that the interaction of Kin1 and Num1 appears to be transient , we cannot rule out that Kin1 and Num1 may interact on microtubule tracks . A possible process that could couple both nuclear and cytoplasmic functions of Num1 is the cytoplasmic transport of mRNAs . In U . maydis , the RNA binding protein Rrm4 was found to be instrumental for bidirectional transport of mRNAs in filaments [47] . Intriguingly , cytoplasmic transport of Rrm4 is dependent on Kin1 , Kin3 and Dynein , and recently it was shown that Rrm4 co-localizes with early endosomes [7] , [77] . There is a remarkable overlap of the phenotypes of rrm4- , num1- as well as kin1-deletion mutants with respect to polarized growth [14] , [78] , which raises the possibility of a concerted function of all three proteins within the same process . Indeed , the motility of early endosomes associated with Rrm4:eGFP fusion proteins is drastically reduced in Δkin1-mutants [7] , [14] and , as we show here , also in num1-deletion strains . In Drosophila oocytes , it has been shown that components of the exon junction complex ( EJC ) are essential for transport of oskar transcripts [79] . The EJC marks exon/exon-transitions during splicing , remains bound to the processed mRNA and serves as a quality control mechanism for the splicing reaction during translation [reviewed in 80] . Reminiscent to mRNA transport in U . maydis , transport of oskar-mRNA is also dependent on the conventional Kinesin 1 [81] . Unlike the components of the EJC that remain bound to mRNA during nuclear export , the NTC is thought to disassemble after the splicing reaction . However , recently it has been shown in S . cerevisiae that the NTC is required for efficient recruitment of TREX to the transcriptional machinery . This is mediated by the C-terminus of the NTC-component Syf1p , which connects TREX to Rpb1 , the largest subunit of RNA-polymerase II [82] . TREX is a conserved protein complex ( “transcription export” ) , which facilitates both transcription elongation as well as mRNA-export processes [83] . Yra1p , a component of the TREX complex , is an RNA-binding protein and interacts with the Mex67/Mtr2-receptor complex , which in turn act together with components of the nuclear pore complex ( NPC ) to guide the transcripts into the cytoplasm [84]–[86] . Thus , it appears well possible that NTC components remain tethered to the spliced mRNA . In this context it is interesting to note the recently discovered association of Kinesin 1- and Dynein to fungal nuclear pores , which mediate NPC-motility and are thus required for chromosome organization and nucleo-cytoplasmic transport [76] . This result corroborates a functional coupling between the nuclear transcription machinery and cytoplasmic transport processes in U . maydis . The major finding of our study is that a component of the conserved spliceosome-associated NTC is involved in cytoplasmic trafficking processes . Due to the interaction of Num1 with the Kin1 and the functional connection to the RNA-binding protein Rrm4 , it is tempting to speculate that Num1 could have a function during intracellular mRNA-transport processes . We hypothesize that one of the functions of the Num1-protein is to participate in the coordination of pre-mRNA splicing and the formation and export of mRNP-particles out of the nucleus . The transcripts would be transferred into the cytoplasm where they could be passed on to the microtubule-based transport machinery via the interaction of Num1 with Kin1 motor proteins that were found to be associated with the nuclear pore complex ( Figure 8 ) . In this model , additional adapter-proteins that act as connecting proteins between the nuclear splicing machinery and the cytoplasmic NPC-bound motor proteins cannot be excluded and remain to be identified .
Escherichia coli strain TOP10 ( Invitrogen ) was used for cloning purposes . Growth conditions and media for the cultivation of E . coli followed the protocols described previously [87] . The Saccharomyces cerevisiae strain AH109 ( Clontech ) was used for yeast two-hybrid interaction studies . S . cerevisiae cells were grown in YPDA complete medium , or on minimal medium ( SD ) supplemented with the dropout-mix needed for selection , as described in the Clontech Matchmaker™ GAL4 Two-Hybrid System 3 Manual ( http://www . clontech . com ) . Ustilago maydis cells were grown in YEPSL [88] , CM ( complete medium ) supplemented with 1% glucose ( CM-G ) and 1% arabinose ( CM-A ) , respectively , or NM ( nitrate minimal medium ) [89] at 28°C . Solid media contained 2% agar . The induction of hyphal growth in AB31- and AB33 derivatives was done as previously described in [24] . Plate mating assays followed the protocol of [24] . Mating assays in liquid culture for the generation of dikaryotic hyphae were carried out as described in [90] . Pheromone stimulation of U . maydis cells was performed according to the protocol of [91] . Aspergillus nidulans was grown in CM ( complex medium ) [92] or MM ( minimal medium ) [93] . All U . maydis and A . nidulans strains used in this study are listed in Tables S7 and S8 , respectively . Plant infections were performed as described in [94] . For infection studies with U . maydis , the maize cultivar Early Golden Bantam was used under controlled conditions in a CLF Plant Climatics GroBank with a 16 h ( 28°C ) /8 h ( 22°C ) day/night rhythm . Disease symptoms were evaluated according to the disease rating criteria published previously [23] . Screening for Num1-interacting proteins was performed as previously described [95] , using the Matchmaker III system ( Clontech ) . Plasmid pGBKT7-Num1 was generated by PCR amplification of the num1 open reading frame ( ORF ) from genomic DNA ( strain 521 [23] ) , introducing two incompatible SfiI restriction sites at the 3′ and 5′ ends and subsequent ligation into the respective sites of pBGKT7 ( Clontech ) . Plasmids pGAD-Prp19 and pGAD-Cef1 containing full-length as well as N- and C-terminal variants were generated by PCR amplification using either genomic DNA ( for cef1 ) or a full-length cDNA library ( for prp19 ) and subsequent ligation into pGAD-DS ( Dualsystems Biotech ) via SfiI sites . Oligonucleotide sequences are given in Table S9 . Molecular techniques followed the protocols described in [87] . DNA from U . maydis was isolated according to the protocol given in [96] . Transformation procedures were performed as described in [97] . DNA from A . nidulans was isolated according to [98] and transformations were carried out as described in [99] . For gene deletions in U . maydis , a PCR-based approach [100] was used . Open reading frames were replaced by either hygromycin , carboxin or nourseothricin resistance cassettes . Similarly , for gene deletions in A . nidulans , a pyridoxine auxotrophic marker was used . C-terminal eGFP , RFP or 3×HA gene fusion constructs were generated in U . maydis via homologous recombination following the protocol given in [101] , using SfiI cassettes from plasmids pUMA647 ( eGFP ) [7] , pUMA738 [47] ( RFP ) or pUMa791 or pUMA792 ( Feldbrügge , unpublished ) for 3×HA fusions . Fusion constructs were subcloned in pCRII TOPO ( Invitrogen ) , PCR-generated linear DNA was used for transformation of U . maydis . Homologous integration of the constructs was verified by Southern analysis . All sequences of oligonucleotides used for PCR are listed in Table S9 . Total RNA was extracted using Trizol reagent ( Invitrogen ) according to the manufacturer's instructions . RNA samples to be used for real-time qRT-PCR or mRNA-Seq analysis were column purified ( RNeasy; Qiagen ) and the quality was checked using a Bioanalyzer with an RNA 6000 Nano LabChip kit ( Agilent ) . Quantitative RT-PCR analysis for pcna , rpb3 , ubi1 , rho3 and rbf1 was conducted with oligonucleotide pairs that specifically detected spliced and unspliced RNA species , respectively: forward primers used for quantification of spliced transcript were specific for exon/intron borders , whereas forward primers for the unspliced transcripts were placed within the intron sequence . Ratios of spliced vs . unspliced RNA of AB31 ( wild-type ) samples were set to 1 . 0 and compared to the ratios obtained with the AB31Δnum1 . qRT-PCR experiments were done in three independent biological and two technical replicates each and followed the protocols given in [102] and [103] . DNA content was measured by flow cytometry . FACS analyses were performed according to [104] with the exception that 105 cells were harvested for each time point and analyzed with a BD Biosciences FACSVerse flow cytometer using the FACSDiva 6 . 2 software . For each acquisition , 2*104 events were measured at an average flow rate of 100–500 events per second . mRNA was sequenced starting from total RNA isolated from strains AB31 and AB31Δnum1 on an Illumina HiSeq 1000 using the TrueSeq mRNA sequencing kit according to the manufacturer's instructions . Three independent libraries for each of the two strains were generated ( Olivier Armant , Insitute for Toxology and Genetics , KIT Campus North ) , using three independently grown cultures ( biological replicates ) . Each library yielded an average of 60 million paired-end reads of 2×56 bp , which equals about 20×109 bp sequences for each strain . Raw paired-end reads ( 2×56 bp ) were aligned with TopHat 1 . 3 . 2 [105] to the Ustilago genome assembly provided by the MIPS Ustilago maydis Database ( MUMDB , http://mips . helmholtz-muenchen . de/genre/proj/ustilago ) . TopHat was supplied with a gtf file derived from MUMDB's p3_t237631_Ust_maydi_20110629 . gtf , but not restricted to known transcripts . Aligned reads were counted using custom Python scripts using the pysam library ( http://code . google . com/p/pysam/ ) . Gene expression values were derived by counting read pairs that were completely inside the exon regions of MUMDB-genes and differential expression was assessed with DESeq 1 . 6 . 1 [106] and BaySeq 1 . 8 . 1 [107] and filtered for an adjusted p-value≤0 . 05 and a more than two fold change . For comparison and visualization of splicing efficiency , read pairs overlapping the respective intron/exon with their first segment by at least three base pairs were counted and normalized to fragments per kb per million ( FPKM ) . Intron retention was determined by calculating the ratio of introns FPKM to the gene's overall exon FPKM . Exon FPKM included all exons with more than 10 bp . Only genes with an FPKM above 10 were considered . A small number of introns with a retention rate above 0 . 9 were excluded , based on the assumption that they base on false gene models . Sequencing data was deposited at the EBI ArrayExpress Database ( E-MTAB-1300 ) . For in vitro protein expression and co-immunoprecipitation of Num1 and Kin1 , the TNT quick coupled transcription/translation system ( Promega ) was used according to the manufacturer's protocol . For expression of Myc-tagged Num1 protein , plasmid pGBKT7-Num1 was used and HA-tagged Kin1 protein was generated using the pGAD-Kin1650–968 derivative identified in the yeast two-hybrid screen . All subsequent steps were carried out as described previously [103] . For in vivo co-immunoprecipitation , U . maydis sporidia were grown in 100 ml CM–G to an OD600 of 0 . 8 . Cells were washed once with PBS buffer , resuspended in 1 ml PBS supplemented with Complete proteinase inhibitor cocktail ( Roche ) , frozen in liquid nitrogen and homogenized in a Retsch mill for 10 minutes at 30 Hz . The cell lysate was incubated with 40 µg monoclonal anti-HA coupled agarose ( Sigma-Aldrich ) on a rotating wheel at 4°C over night . Agarose beads were washed three times in PBS prior to resuspension in 20 µl Laemmli buffer . Samples were boiled for three minutes and separated by SDS-PAGE . For co-immunoprecipitation of Num1 and Kin1 , a crosslinking reaction with dithiobis[succinimidylpropionate] ( Thermo Fisher ) was conducted according to [108] using 300 ml of cells grown in CM-G to an OD600 of 0 . 8 . Proteins were transferred to PVDF nitrocellulose membranes in a semidry blot chamber . Western blots were probed with monoclonal anti-HA , anti-c-Myc , and anti-GFP ( Sigma-Aldrich ) antibodies . Horseradish peroxidase–conjugated anti-mouse or anti-rabbit IgG ( Promega ) was used as secondary antibody , and an ECL system was used for protein detection . For microscopic analyses , logarithmically growing U . maydis cells were taken from liquid cultures grown in CM-G medium . For the induction of hyphal growth , cells were shifted to CM-A or NM to induce the Pcrg1- or Pnar1- promoters , respectively , for 12–14 hours . For microscopy of A . nidulans germlings and young hyphae , MM on cover slips was inoculated with a small amount of spores and incubated for 12–18 hours at 37°C . Cells were then placed on top of a 2% agarose cushion placed on the microscope slide and immediately observed using an Axioimager Z1 microscope equipped with an Axiocam MRm camera ( Carl Zeiss ) . Standard filter sets for DAPI , GFP , CFP and Rhodamine were used for epifluorescence analysis . Nuclei were stained with DAPI Vectashield H-1200 ( Vector Laboratories ) ; chitin was stained with 2 µg/ml Calcofluor White ( Sigma-Aldrich ) or with Congo Red solution ( 1 µl/ml; stock: 1 mg/ml in H2O ) . Congo Red was incubated at 25°C for 10 minutes on a turning wheel and washed twice with CM prior to microscopic analysis . Fungal cell walls in general were stained with 1 mg/ml WGA/Fluorescein ( Invitrogen ) in PBS . For the visualization of vacuoles , growing cells were incubated in CM supplemented with 10 µg/ml CellTracker Blue ( 7-amino-4-chloromethyl-coumarin , CMAC ) ( Invitrogen ) for 30 minutes at 28°C , washed twice with CM and then subjected to microscopy . Chlorazole Black E staining was performed according to [24] . Endosome motility was measured in image sequences of 50 frames , taken with an exposure time of 500 msec . Only organelles that were moving for a distance of at least 5 µm were considered . The resulting movies were converted into kymographs using ImageJ software [109] . For quantification of Mcs1 signal intensity , the corrected total cell fluorescence ( CTCF ) of mid-size growing buds was calculated with ImageJ software ( CTCF = integrated density – ( area of selected cell×fluorescence of background reading ) ) . For quantification of the microtubule orientation , cells were grown to mid-log phase in YEPSL medium . Cells were analyzed via 45 sec time-lapse recordings ( four frames each 15 sec ) . Pictures were taken with an exposure time of 90 msec for Tub1-GFP and 300 msec for Peb1:RFP signals . Movement of Peb1:RFP signals was scored at a screen as described in [46] . All image processing , including adjustment of brightness , contrast and gamma-values was performed with the AxioVision and ZEN software ( Carl Zeiss ) , respectively . For comparative phylogenetic analysis of Num1 , 65 sequences with the highest similarity to Num1 were obtained by BLASTP analysis . The sequence of Saccharomyces cerevisiae Snt309p was included as outgroup . Sequences were aligned with MAFFT version 6 [110] using the global alignment G-INS-i . A phylogenetic tree was calculated using the minimum linkage clustering method ( http://align . bmr . kyushu-u . ac . jp/mafft/online/server/ ) . Fig Tree 1 . 3 ( http://tree . bio . ed . ac . uk/software/figtree/ ) was used to visualize the Nexus formats of the MAFFT results . Sequence data from this article can be found at the Munich Information Center for Protein Sequences ( MIPS ) Ustilago maydis database ( http://mips . helmholtz-muenchen . de/genre/proj/ustilago/ ) , the Aspergillus genome database ( http://www . aspgd . org/ ) and the National Center for Biotechnology Information ( NCBI ) database under the following accession numbers: num1 ( um01682 ) , XP_757829; prp19 ( um10027 ) , XP_756259; cef1 ( um04411 ) , XP_760558; rbf1 ( um03172 ) , XP_759319; pcna ( um05403 ) , XP_761550; rpb3 ( um03550 ) , XP_759697; ubi1 ( um02440 ) , XP_758587; rho3 ( um04070 ) , XP_760217; rad51 ( um03290 ) , XP_759437; rec1 ( um11016 ) , XP_759527 . 1; mms2 ( um10097 ) , XP_756467 . 1; cks1 ( um03210 ) , XP_759357; cdk1 ( um10705 ) , AAP94021 . 1; hdp1 ( um12024 ) , XP_761909 . 1; hdp2 ( um04928 ) , XP_761075; biz1 ( um02549 ) , XP_758696; um02704 ( related to allantoate permease ) , XP_758851; um12105 ( probable PUP3 – 20S proteasome subunit ) , XP_756658 . 1; actin ( um11232 ) , XP_762364; eIF2b ( um04869 ) , XP_761016; kin1 ( um04218 ) , XP_760356; dyn2 ( um04372 ) , XP_760519; yup1 ( um05406 ) , XP_761553; tub1 ( um01221 ) , XP_757368; AN4244 , XP_661848 . 1; nudA ( AN0118 ) , XP_657722 . 1; kinA ( AN5343 ) , XP_662947 . 1 . ; the gene model for um15049 ( related to PUF3 transcript specific regulator of mRNA degradation ) is implemented only at MUMDB ( http://mips . helmholtz-muenchen . de/genre/proj/ustilago ) , as the predicted protein is composed of two fragments located on adjacent contig ends and , according to the sequence gap , part of the protein may be absent . | In eukaryotic cells , nascent mRNA is processed by splicing to remove introns and to join the exon sequences . The processed mRNA is then transported out of the nucleus and employed by ribosomes to synthesize proteins . Splicing is achieved by the spliceosome and associated protein complexes , among them the so-called NineTeen complex ( NTC ) . We have identified the Num1 protein as one of the core components of the NTC in the fungus Ustilago maydis , and could show that it is required for polarized growth of the filamentous fungal cells . Consistent with the NTC function , cells with a num1-deletion show reduced splicing of mRNA . Moreover , we uncover a novel cytoplasmic function of the Num1 protein: It physically interacts with the microtubule-associated Kinesin 1 motor protein , and phenotypic analyses corroborate that both proteins are functionally connected . Our findings reveal a yet unidentified role of a global splicing factor during intracellular trafficking processes . A possible connection between these disparate mechanisms presumably resides in mRNA-export out of the nucleus and/or the transport of mRNA within the cytoplasm . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2014 | The SPF27 Homologue Num1 Connects Splicing and Kinesin 1-Dependent Cytoplasmic Trafficking in Ustilago maydis |
Systemic acquired resistance , mediated by the Arabidopsis NPR1 gene and the rice NH1 gene , confers broad-spectrum immunity to diverse pathogens . NPR1 and NH1 interact with TGA transcription factors to activate downstream defense genes . Despite the importance of this defense response , the signaling components downstream of NPR1/NH1 and TGA proteins are poorly defined . Here we report the identification of a rice mutant , snim1 , which suppresses NH1-mediated immunity and demonstrate that two genes encoding previously uncharacterized cysteine-rich-receptor-like kinases ( CRK6 and CRK10 ) , complement the snim1 mutant phenotype . Silencing of CRK6 and CRK10 genes individually in the parental genetic background recreates the snim1 phenotype . We identified a rice mutant in the Kitaake genetic background with a frameshift mutation in crk10; this mutant also displays a compromised immune response highlighting the important role of crk10 . We also show that elevated levels of NH1 expression lead to enhanced CRK10 expression and that the rice TGA2 . 1 protein binds to the CRK10 promoter . These experiments demonstrate a requirement for CRKs in NH1-mediated immunity and establish a molecular link between NH1 and induction of CRK10 expression .
Despite the lack of circulating immune cells , plants share similarities with animals in their defense responses against pathogens; they both use innate immune systems to counter attacks . For example , plants and animals utilize membrane-localized receptors that detect conserved molecular patterns derived from microbes [1] , including peptidoglycan , flagellin , chitin , and sulfated peptides . Another component of the immune system consists of cytoplasmic nucleotide-binding-domain , leucine-rich repeat receptors ( NLR ) [2] that recognize cognate effector proteins secreted by the microbe into the cell . Plants have a third type of defense system , absent in animals , called systemic acquired resistance ( SAR ) [3 , 4] . Activation of this system leads to a long lasting , broad-spectrum defense response . SAR is induced by the plant hormone salicylic acid ( SA ) , and its analogues 2 , 6-dichloroisonicotinic acid ( INA ) , probenazole , and benzothiadiazole ( BTH ) [5–10] . In Arabidopsis , NPR1 ( nonexpressor of pathogenesis-related genes 1; also known as NIM1 and SAI1 ) is the key regulator of SAR [11–15]; SA induces NPR1 expression and activates the NPR1 protein leading to immunity against diverse pathogens [16–19] . NPR1 interacts with TGA transcription factors , which are required for the SAR response [20–24] . NPR1 contains an ankyrin-repeat domain , a BTB/POZ domain , and a C-terminal transcription activation domain . The ankyrin-repeat domain interacts with TGA proteins [20]; the BTB/POZ domain regulates the C-terminal activation domain [25 , 26] . SA binds NPR1 and may regulate NPR1 directly [27]; another model suggests that SA regulates NPR1 indirectly through NPR3 and NPR4 , which bind SA with high affinities [28] . Arabidopsis NIMIN ( NIM1 interacting ) and rice NRR ( negative regulator of resistance ) family proteins are the second class of proteins that interact with Arabidopsis NPR1/NIM1 and rice NH1 ( NPR1 homolog 1; also known as OsNPR1 ) [29 , 30] , respectively . Arabidopsis NIMINs [31] and rice NRR members ( NRR , RH1 , RH2 , and RH3 ) [32] negatively regulate NPR1 and NH1 , respectively [31–33] . In rice , elevated expression of Arabidopsis NPR1 or the rice ortholog NH1 or the NH1 paralog NH3 [22 , 29 , 30 , 34] results in enhanced resistance to Xanthomonas oryzae pv . oryzae ( Xoo ) and Magnaporthe oryzae , the causal agents of rice bacterial leaf blight and rice blast , respectively . Like Arabidopsis NPR1 , rice NH1 also interacts with TGA transcription factors [29] . The enhanced disease resistance of NH1 overexpression ( NH1ox ) rice plants is accompanied by and correlated with a spontaneous cell death phenotype , commonly referred to as a lesion mimic phenotype [29 , 35 , 36] . Application of BTH enhances the formation of necrotic spots from the spontaneous cell death on the NH1ox plants [29] , indicating a tight association between this lesion mimic phenotype and enhanced resistance to Xoo . Although Arabidopsis NPR1 and rice NH1 bind to TGA proteins and act as transcriptional co-activators [26 , 37] , the downstream components required for the NPR1/NH1-mediated response remain largely uncharacterized . To identify such proteins , we screened a fast-neutron-induced rice mutant population of NH1ox plants treated with BTH and identified a mutant called suppressor of NH1-mediated immunity 1 ( snim1 ) that no longer responds to BTH . Here , we report the identification of two previously uncharacterized cysteine-rich receptor-like kinases ( encoded by CRK6 and CRK10 ) that complement the snim1 phenotype and are required for the BTH-induced immune response . We demonstrate that elevated NH1 levels induce CRK10 expression and that the rice TGA2 . 1 protein binds to the CRK10 promoter , indicating that both NH1 and TGA proteins regulate CRK10 expression .
In our screen for suppressors of NH1-mediated immunity [38] , we isolated the snim1 mutant . After inoculation with Xanthomonas oryzae pv . oryzae ( Xoo ) , snim1 develops long water-soaked lesions , characteristic of the disease . In contrast , the parental control NH1ox plants are resistant to infection , resulting in short lesions ( Fig 1A ) . Furthermore , the NH1ox parent displays a lesion mimic spontaneous cell death phenotype after application of BTH , which is a plant defense activator [38] , but snim1 lacks this phenotype ( Fig 1B ) . Bacterial population measurements show that snim1 harbors 18 times more Xoo than NH1ox plants ( P = 0 . 0052 ) ( Fig 1C ) . These results indicate that the snim1 mutation impairs NH1-mediated immunity . To expedite isolation of the gene ( s ) responsible for the snim1 phenotype , we conducted comparative genome hybridization ( CGH ) analysis using a NimbleGen 2 . 1-million-probe rice tiling array [38] comparing snim1 and NH1ox DNA ( Fig 2A ) . These experiments identified a single large deletion ( from approximately 21 , 297 , 000 to 21 , 385 , 000 ) on chromosome 7 of snim1 ( Fig 2B ) . This 88kb deletion was confirmed by PCR . The deleted region contains 11 genes: six CRKs , one encoding development and cell death-kelch motif protein , three encoding expressed proteins , and one Ty3-gypsy retrotransposon gene . Their MSU gene IDs and relative positions are shown in S1 Fig . Analysis of F2 progeny derived from a cross between snim1 and the parent reveals a complete association of the 88kb deletion with the snim1 phenotype ( Fig 3 ) . F2 progeny that contain the 88kb region exhibit levels of Xoo resistance similar to that of the NH1ox parent ( Fig 3 ) . In contrast , F2 progeny lacking the 88kb region are susceptible to Xoo . T-test yields P<0 . 0001 , indicating the difference between the two groups is highly significant . To assess the involvement of the 10 non-retrotransposon genes deleted in snim1 , we conducted complementation experiments . We ligated each isolated gene into the binary vector C4300 , and used the resulting construct to transform snim1 . Following inoculation with Xoo , we found that only the CRK6 and CRK10 constructs restore resistance to the snim1 mutant ( Fig 4 ) . None of the other 8 genes restore resistance to Xoo ( S2 Fig ) . To investigate the relationship between the rice CRKs and their Arabidopsis paralogs , we constructed a phylogenetic tree for the 45 CRKs and included the Arabidopsis CRKs to display their relative positions ( S3 Fig ) . To further validate the requirement of CRK6 and CRK10 for NH1-mediated immunity to Xoo , we silenced CRK6 and CRK10 individually in the NH1ox background using RNA interference ( Ri ) and inoculated the resulting six transgenic lines with Xoo . Real-time quantitative reverse transcription ( qRT ) -PCR experiments revealed that the CRK6Ri and CRK10Ri lines are effectively silenced for CRK6 and CRK10 expression ( 70%-90% reduction ) ( S4 Fig ) . Accumulation of CRK6 and CRK g35580 RNAs are also reduced in the CRK10Ri lines . In contrast , RNA levels of CRK g35650 , g35660 and g35680 are not reduced in the CRK10Ri lines . These results indicate that silencing of CRK10 also affects CRK6 and CRK g35580 expression levels . Based on the low sequence similarity of CRK6 and CRK g35580 with the region of CRK10 used for silencing , we hypothesize that the observed reduced expression is indirect and not due to co-silencing . In contrast to the resistant NH1ox parent plants , which display spontaneous cell death , lesion mimic spots following BTH treatment ( Fig 5A ) and are resistant to Xoo ( Fig 5B ) , the CRK6Ri and CRK10Ri lines lack the spontaneous cell death phenotype ( Fig 5A ) and are susceptible to Xoo ( Fig 5B and 5C ) . The presence of the CRK10Ri and CRK6Ri transgenes cosegregates with susceptibility to Xoo in T1 progeny ( S5 and S6 Figs ) . Bacterial growth curve analyses reveal that CRK6Ri ( lines #3&10 ) and CRK10Ri ( #4&13 ) plants harbor 5–8 times and 11–12 times more Xoo than the NH1ox parent , respectively ( Fig 5D ) . Statistical analysis of bacterial populations at day 12 ( Fig 5D ) indicated significant differences between NH1ox parent , CRK6Ri , and CRK10Ri lines . These results confirm the requirement of CRK6 and CRK10 for NH1-mediated immunity . We next tested whether the snim1 mutation affects resistance to Xoo in the absence of the NH1ox transgene . For this purpose , we crossed snim1 with the parental rice cultivar LiaoGeng ( LG ) , which lacks the NH1ox transgene . We identified 27 individual progeny ( snim1/LG ) from this cross that carry homozygous snim1 88kb deletion and lack the NH1ox transgene . We pretreated the snim1/LG progeny and the LG control with 1mM BTH and then inoculated with Xoo . The snim1/LG individuals display significantly ( P<0 . 0001 ) longer lesions ( 9 . 0±1 . 9 cm ) than the LG control ( 6 . 0±1 . 5 cm ) ( Fig 6A ) . These results indicate that the snim1 mutation also affects BTH-induced resistance to Xoo in the absence of the NH1ox transgene . We further verified the role of the CRK10 gene with a crk10 rice mutant . We identified a homozygous crk10 mutant from a fast neutron mutagenized rice population in the Kitaake genetic background [39] . This mutant carries a two-base deletion ( S7 Fig ) in the third exon of CRK10 causing a frameshift , resulting in a predicted truncated CRK10 protein missing the kinase domain ( see Methods ) . To assess the effect of the CRK10 mutation in this second genetic background , we inoculated 27 progeny plants of this crk10 mutant along with 25 wild type Kitaake control plants with Xoo following 1mM BTH treatment . We observed that the crk10/Kitaake mutant plants develop significantly ( P = 0 . 0009 ) longer lesions ( ~12 . 5 cm ) than the control Kitaake plants ( ~10 . 5 cm ) ( Fig 6B ) . These results further support the conclusion that the CRK10 gene is required for resistance to Xoo . To test the hypothesis that higher levels of CRK10 enhance resistance , we generated an inducible construct ( GVG-CRK10 ) using the GVG-dexamethasone ( DEX ) inducible system [40 , 41] and obtained 10 healthy , independently transformed lines . These plants display enhanced resistance to Xoo after DEX induction ( Fig 7A–7C ) . The enhanced resistance cosegregates with the GVG-CRK10 transgene in T1 segregating progeny ( S8 Fig ) . Further analysis indicates that higher levels of CRK10 lead to higher levels of resistance to Xoo ( Fig 7D ) . Overexpression of CRK6 did not confer enhanced resistance to Xoo ( S9 Fig ) . To determine if NH1 regulates expression of CRK genes , we assessed CRK10 and CRK6 expression levels in Kitaake and NH1 overexpression plants ( nNH1 in Kitaake genetic background ) . We treated plants with 1 mM BTH and collected leaf samples 0 , 1 , 4 , 8 , 24 , and 48 hours after treatment . BTH treatment induces NH1 expression levels in both Kitaake ( labeled Kit ) and nNH1 plants , reaching peaks at 8 hours ( Fig 8A ) . BTH treatment induces 3-fold higher levels of CRK10 mRNA ( at 8 hr ) in Kitaake compared with the untreated control ( 0 hr ) ; this induction is higher ( 4 . 5-fold ) in the nNH1 plants ( at 8 hr; Fig 8A ) . In the absence of BTH pretreatment , the CRK10 mRNA level is 2 . 5-fold higher in nNH1 plants compared with control plants . BTH treatment also induces CRK6 expression ( by 4 fold at 8hr ) . We observed a slight delay in CRK6 induction in nNH1 plants . This delay does not significantly change the basal and peak levels of CRK6 expression ( S10 Fig ) . To further investigate the observation that NH1 regulates CRK10 expression , we assessed CRK10 expression levels in transgenic Kitaake plants carrying an NH1 RNAi construct ( NH1Ri ) , under control of the DEX-inducible promoter . The NH1 mRNA levels are significantly reduced in the NH1Ri plants ( NH1Ri-6A and -7A ) silenced for expression of NH1 compared with the Kitaake control , 6 and 8 hours after DEX and BTH treatment ( Fig 8B ) . Importantly , we also observed that CRK10 mRNA levels are also significantly reduced in the NH1Ri plants compared with the Kitaake control ( Fig 8B ) . These results indicate that BTH and NH1 regulate CRK10 expression . To assess the mechanism by which CRK10 expression is activated , we examined the CRK10 promoter for potential transcription factor binding sites . We identified the sequences TGACGT ( -793 from the TATA box ) and TGACG ( -164 ) in the CRK10 and CRK6 promoters , respectively , that match the consensus sequence binding sites for TGA transcription factors . We synthesized an oligonucleotide containing this putative CRK10 TGA binding site and employed an electrophoresis mobility shift assay ( EMSA ) to test its interaction with the rice TGA2 . 1 protein , which has been shown to bind to an SA-responsive element and interact with NH1 [42 , 43] . We found that rice TGA2 . 1 binds to the synthetic oligonucleotide and that the observed gel shift can be competed with higher concentrations of wild-type oligonucleotides but not by oligonucleotides carrying a mutation in the TGA binding site ( Fig 9A ) . These results demonstrate that the CRK10 promoter contains a cognate TGA binding site . CRK6 and CRK10 carry conserved kinase motifs . To assess the kinase activity of CRK6 , we fused the CRK6 kinase domain ( CRK6K ) to the His:Nus protein , expressed the fusion construct in E . coli and purified the fusion protein using Ni-NTA resins . As a negative control , we generated a kinase-dead mutant of the CRK6 kinase by mutating the conserved , required aspartate at amino acid 488 to asparagine ( CRK6DN ) . The two fusion proteins were subject to kinase activity assay ( see Methods ) . The Nus:CRK6K protein shows clear autophosphorylation , whereas the Nus:CRK6DN protein did not ( S11A Fig ) . These results indicate that CRK6 is an active kinase . We were unable to express and purify the CRK10 kinase . However , the kinase domains of CRK6 and CRK10 are highly conserved sharing 76% similarity ( S11B Fig ) , including all amino acids known to be critical for kinase activity suggesting that CRK10 is likely also an active kinase . In addition to the kinase domain , we have also analyzed the CRK6 and CRK10 proteins for other conserved protein domains with the SMART program , which is specialized in detecting protein domains . Both CRK6 ( amino acids 1–31 ) and CRK10 ( aa . 1–27 ) contain a predicted signal peptide . Both also contain a predicted transmembrane region: amino acids 295–317 for CRK6 and 284–306 for CRK10 . These results strongly predict that CRK6 and CRK10 are membrane-localized proteins .
In this manuscript we demonstrate that the previously uncharacterized proteins , CRK6 and CRK10 , are required for BTH-inducible , NH1-mediated immunity in rice . There are 43 additional CRK genes in rice ( http://rice . plantbiology . msu . edu ) and at least 44 CRK members in Arabidopsis [44 , 45] . In both rice and Arabidopsis , many of the CRK genes are clustered together in the genome . This structure may facilitate recombination and accelerate evolution of resistance . A similar mechanism for generating diversity has been postulated for plant resistance genes encoding both nucleotide-binding site leucine-rich repeats proteins and leucine rich repeat receptor kinases [46 , 47] . Although altered expression of CRK genes has been observed in several datasets in response to biotic stress [48–53] , the biological functions of CRK genes have not previously been well elucidated for any of the Arabidopsis or rice CRK genes for their involvement in the immune response . In particular , the requirement of these genes in NPR1- or NH1-mediated immunity has not previously been demonstrated . The observed functional redundancy of CRK genes in monocots and dicots has likely hindered their characterization and prevented an unambiguous assignment of their function [53] . Because of these complications , previous analyses of CRK proteins relied almost exclusively on overexpression experiments . For example , overexpression of AtCRK4 , AtCRK5 , AtCRK6 , AtCRK13 , and AtCRK45 resulted in cell death , activation of defense genes , and/or increased resistance to Pseudomonas syringae pv . tomato DC3000 . Of the knockout analyses conducted , only the Atcrk45 mutant displayed a slight alteration in response to Pst DC3000 [50] . Here we provide direct and robust genetic evidence that the CRK6 and CRK10 genes mediate BTH-induced immune response . We have demonstrated that the snim1 mutant and the crk10 knockout mutant compromise resistance to Xoo in two genetic backgrounds in the absence of NH1 overexpression ( Fig 6 ) . Furthermore , CRK10 is likely broadly involved in resistance to pathogens in a quantitative manner because higher levels of CRK10 correlate with enhanced cell death and immune response ( Fig 7 ) . This result is consistent with the observations that CRK genes are induced by diverse pathogens [48–53] . CRK10 may play a more significant role in the immune response than other rice CRK genes because our results clearly show that silencing ( Fig 5 ) and knockout ( Fig 6B ) of CRK10 itself both cause obvious phenotypes in the immune response; CRK6-silenced lines show a less susceptible phenotype . Our genetic data reveal that elevated expression of NH1 results in enhanced CRK10 expression . Conversely , a reduction in NH1 expression leads to a reduction of CRK10 expression . These results unambiguously demonstrate that CRK10 expression is regulated by the NH1 protein . The observations that TGA proteins bind to the CRK10 promoter ( Fig 9A ) and interact with NH1 [43] suggest that NH1 and TGA proteins function together to activate expression of CRK10 . Fig 9B presents a model for NH1 and CRK10-mediated activation of defense responses . Based on the observation that NPR1 or NPR3/NPR4 proteins bind SA [27 , 28] , we hypothesize that rice NH1 also senses BTH/SA and that the activated NH1 protein translocates to the nucleus where it interacts with TGA transcription factors . The TGA/NH1 protein complex then activates expression of downstream genes , including CRK6/CRK10 .
The rice mutant population in the NH1ox-54 genetic background ( in the rice LiaoGeng variety carrying the Ubi-NH1 gene ) was generated by irradiation with fast neutrons at 20 Grays as previously described [38] . Xoo inoculation was carried out in a growth chamber , set at 26°C with 80% humidity . Inoculation of rice plants with Xoo strain PXO99 was carried out with the scissor-dip method [54] with absorbance ( 600 nm ) at OD = 0 . 5 . The crk10/Kitaake mutant was generated in the Xa21 background and was inoculated with the ΔraxST/PXO99 strain , which evades Xa21-mediated immunity [55] . Comparative genome hybridization was carried out at the Roche NimbleGen facility ( Madison , WI ) using the Roche NimbleGen rice whole genome tiling array as described before [38] . A Qiagen Long Range PCR kit was used for amplification of each gene including the promoter , the coding region , and the 3’ sequence , from chromosome 7 . Amplification of g35580 used primers G580-1 and G580-3 , g35600 used primers G600-1 and G600-2 , g35610 used primers G610-1 and G610-3 , g35620 used primers G620-1a and G620-2a , g35630 used primers G630-1 and G630-2 , g35640 used primers G640-1 and G640-2 , g35650 used primers G650-1 and G650-2 , g35660 used primers G660-1 and G660-2 , g35680 used primers G680-1 and G680-3 , g35690 used primers G690-1 and G690-3 , and g35700 used primers G700-1 and G700-2 . Amplification of g35600 , g35610 , g35640 , g35650 , g35660 , g35690 , g35700 used Liaogeng genomic DNA as the PCR template . The remaining genes used PAC clone P0458H05 as template . PCR products were cloned into the pCR8/GW/TOPO vector ( Invitrogen ) and confirmed by sequencing . Each gene was subcloned into the C4300 vector by Gateway recombination . Genotyping of plants carrying the 88-kb deletion was carried out with primers targeting genes CRK6 , CRK10 , or Os07g35610 . CRK6 genotyping used primers G690-RT1 and G690-RT2 . CRK10 genotyping used primers G700-RT3 and G700-RT4 . Os07g35610 genotyping used primers G610-10 and G610-2 . A two-base deletion in exon 3 of the CRK10 gene ( Os07g35700 ) is present ( line # FN892-S , M2 ) compared to the Kitaake parent that causes a frame shift disrupting the CRK10 open reading frame . M3 progeny plants were used for Xoo inoculation . To generate an RNAi construct targeting CRK6 ( Os07g35690 ) , we used primers G690-SiRI and G690-SiBam to amplify a 500bp fragment from the 5’-end of CRK6 . This fragment was digested with EcoRI and BamHI and cloned into plasmid pENTR/L16 , modified from pENTR/D to contain multiple cloning sites . The clone was confirmed by sequencing . The fragment was excised with EcoRI and BamHI and subcloned into pBluescript II SK- , pre-cut with BamHI and phosphatase-treated , jointly with the Xa21 intron ( precut with EcoRI ) . The resulting clone ( dsG690/SK ) contained two pieces of the CRK6 fragment head-to-head with the Xa21 intron in between to serve as a spacer to stabilize the clone in bacteria . The dsG690 insert was excised with BamHI and subcloned back to the pENTR/L16 vector using the BamHI site and recombined with a Gateway compatible Ubi-C4300 binary vector ( Ubi-C4300/GA ) to yield construct Ubi-dsG690/C4300 . To generate an RNAi construct targeting CRK10 , we used primers G700-SiRI and G700-SiBam to amplify a 500-bp fragment from the 5’-end of CRK10 . The PCR product was processed the same way as the CRK6 fragment for generating the end product Ubi-dsG700/C4300 construct . These constructs were also used to transform the NH1ox-11 line . Genotyping of CRK6Ri plants used primers G690-SiRI and Ubi-1 primers; genotyping of CRK10Ri plants used G700-SiRI and Ubi-1 primers . A full-length , 2 kb CRK10 cDNA was amplified with primers G700-3 and G700-8 from Nipponbare and cloned into pENTR/D . To generate an inducible CRK10 construct in the GVG-DEX system to overexpress CRK10 , the CRK10 cDNA was subcloned into vector TA7002/GA by Gateway recombination , creating construct GVG-G700 . Genotyping of the GVG-G700 construct in transgenic lines used primers Hyg-3 and Hyg-4 , targeting the hygromycin selection marker . For silencing of NH1 , an RNAi construct was generated using the GVG-DEX inducible vector . A 500 bp NH1 5’-cDNA fragment was excised from NH1 cDNA with EcoRI and SalI . This fragment was ligated with a Gus spacer digested with EcoRI into the pENTR/L16 vector . The resulting construct was recombined with a Gateway compatible pTA7002 binary vector to generate RNAi construct GVG-NH1Ri targeting NH1 . Total RNA was extracted using the Trizol reagent ( Invitrogen ) and purified with spin-column ( NucleoBond ) . One to five μg of total RNA each sample was used to synthesize cDNA for real time RT-PCR . To assess the expression level of CRK6 , primers G690-Q1a and G690-Q2 or primers G690-Q1b and G690-Q2 were used . These primers were determined to be specific to the CRK6 gene . For CRK10 , primers G700-RT3 and G700-RT5 were used for real time RT-PCR . BTH was applied to rice leaves in a greenhouse in the form of a foliar spray at a concentration of 1 mM in the form of Actigard ( Syngenta ) . DEX was dissolved in DMSO and diluted to 100 μM in 0 . 05% Tween 20 and applied by foliar spray . For the EMSA assay , a probe was generated via annealing two oligonucleotides containing the putative TGA binding site . The top oligonucleotide contains biotin at the 5’end . Detection of biotin on the probe by streptavidin was carried out using a Chemiluminescent Nucleic Acid Detection Module ( Thermo Scientific , Rockford , IL ) . Statistical analysis was carried out using the JMP Pro 10 statistics program . Protein expression in E . coli BL21 cells and purification of the fusion protein was carried out according to Chen et al . [56] . Kinase activity assay was conducted as described by Chen et al . [56] . | To survive , plants and animals must resist microbial infection . Plants employ an immune response called systemic acquired resistance that confers long-lasting resistance to a broad-spectrum of pathogens . Researchers have previously identified two key proteins ( NPR1/NH1 and TGA ) that control this immune response . Despite these advances , there remain many gaps in our knowledge and understanding of this important immune response . We have identified a new gene ( CRK10 ) required for this immune response; without it , plants are more susceptible to infection . These findings advance basic knowledge of systemic acquired resistance and open the door to a new avenue of research on this exciting and important resistance mechanism . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"biotechnology",
"plant",
"anatomy",
"electrophoretic",
"mobility",
"shift",
"assay",
"gene",
"regulation",
"regulatory",
"proteins",
"rna",
"extraction",
"dna-binding",
"proteins",
"plant",
"physiology",
"cereal",
"crops",
"plant",
"science",
"rice",
"model",
"organisms",
"genetically",
"modified",
"plants",
"transcription",
"factors",
"crops",
"plant",
"pathology",
"molecular",
"biology",
"techniques",
"plants",
"extraction",
"techniques",
"genetic",
"engineering",
"research",
"and",
"analysis",
"methods",
"grasses",
"genetically",
"modified",
"organisms",
"crop",
"science",
"gene",
"mapping",
"proteins",
"gene",
"expression",
"leaves",
"molecular",
"biology",
"agriculture",
"biochemistry",
"plant",
"defenses",
"plant",
"and",
"algal",
"models",
"plant",
"disease",
"resistance",
"agricultural",
"biotechnology",
"genetics",
"protein",
"domains",
"biology",
"and",
"life",
"sciences",
"restriction",
"fragment",
"mapping",
"plant",
"biotechnology",
"organisms"
] | 2016 | A Genetic Screen Identifies a Requirement for Cysteine-Rich–Receptor-Like Kinases in Rice NH1 (OsNPR1)-Mediated Immunity |
Fibrillins are large extracellular macromolecules that polymerize to form the backbone structure of connective tissue microfibrils . Mutations in the gene for fibrillin-1 cause the Marfan syndrome , while mutations in the gene for fibrillin-2 cause Congenital Contractural Arachnodactyly . Both are autosomal dominant disorders , and both disorders affect musculoskeletal tissues . Here we show that Fbn2 null mice ( on a 129/Sv background ) are born with reduced muscle mass , abnormal muscle histology , and signs of activated BMP signaling in skeletal muscle . A delay in Myosin Heavy Chain 8 , a perinatal myosin , was found in Fbn2 null forelimb muscle tissue , consistent with the notion that muscle defects underlie forelimb contractures in these mice . In addition , white fat accumulated in the forelimbs during the early postnatal period . Adult Fbn2 null mice are already known to demonstrate persistent muscle weakness . Here we measured elevated creatine kinase levels in adult Fbn2 null mice , indicating ongoing cycles of muscle injury . On a C57Bl/6 background , Fbn2 null mice showed severe defects in musculature , leading to neonatal death from respiratory failure . These new findings demonstrate that loss of fibrillin-2 results in phenotypes similar to those found in congenital muscular dystrophies and that FBN2 should be considered as a candidate gene for recessive congenital muscular dystrophy . Both in vivo and in vitro evidence associated muscle abnormalities and accumulation of white fat in Fbn2 null mice with abnormally activated BMP signaling . Genetic rescue of reduced muscle mass and accumulation of white fat in Fbn2 null mice was accomplished by deleting a single allele of Bmp7 . In contrast to other reports that activated BMP signaling leads to muscle hypertrophy , our findings demonstrate the exquisite sensitivity of BMP signaling to the fibrillin-2 extracellular environment during early postnatal muscle development . New evidence presented here suggests that fibrillin-2 can sequester BMP complexes in a latent state .
The fibrillins are large structural macromolecules that are important constituents of all connective tissues . Fibrillins form microfibrils [1–3] , and they target growth factors to the extracellular space [4–6] . In humans , there are 3 fibrillin genes , whereas in mouse , the gene for fibrillin-3 was lost due to chromosomal rearrangement [7] . Expression of genes for fibrillin-2 and -3 is limited largely to fetal development [7 , 8] . However , fibrillin-2 protein remains within the tissue microfibrils in postnatal tissues [9] . Mutations in the gene for fibrillin-2 ( FBN2 ) cause congenital contractural arachnodactyly ( CCA ) or Beals syndrome ( now designated as Distal Arthrogryposis , Type 9 , OMIM #121050 ) , a dominantly inherited disorder of connective tissue . Muscle weakness was noted in some cases of CCA [10] , but it has not been emphasized ( OMIM #121050 ) . Characteristic features include contractures of the large and small joints , arachnodactyly , scoliosis , and crumpled ears . In contrast to humans with a heterozygous mutation , Fbn2 null mice are born with syndactyly and contractures [11] . They also have persistent impairment of locomotory function in their hindlimbs which is not due to defects in the peripheral nervous system [12] . Similar muscle weakness has been reported in the Fbn2 mutant “Mariusz” mouse , identified as part of a large-scale N-ethyl-N-nitrosourea ( ENU ) mutagenesis screen [13] . Abnormal activation of TGFβ signaling was found to contribute to several features of the Marfan syndrome ( MFS ) , a dominantly inherited disorder caused by mutations in the gene for fibrillin-1 ( FBN1 ) . Hallmark features of MFS include aortic aneurysm and dissection; tall stature , arachnodactyly , scoliosis , and loose joints; and ectopia lentis . In addition , pneumothorax and skeletal muscle hypoplasia are also characteristic of MFS . Activated TGFβ signaling was found in the lung [14] , aorta [15] , and skeletal muscle [16] of Marfan mice . Loss of fibrillin-2 in mice was associated with syndactyly and loss of BMP signaling during limb bud patterning [11] . Dysregulated growth factor signaling has not been found in musculoskeletal tissues associated with contractures or hypermobility in CCA or MFS . Scoliosis and arachnodactyly are features of both CCA and MFS . However , instead of the hypermobile joints characteristic of MFS , there are contractures of the large and small joints in CCA . Certain FBN1 mutations cause syndromes with musculoskeletal features that are the opposite of MFS ( short stature acromelic dysplasias such as Weill-Marchesani syndrome ( WMS ) [17 , 18] and geleophysic dysplasia or acromicric dysplasia [19] ) . Skeletal muscle mass in MFS is decreased , whereas hypermuscularity is a feature of WMS . The conundrum posed by opposing clinical features resulting from mutations in the same gene ( FBN1 ) or in related genes ( FBN1 or FBN2 ) may be explained by differential effects of mutations on the activation or inhibition of growth factor signaling . Although the genetic data demonstrate differential , or in some cases even opposite , phenotypic effects of fibrillin mutations , knowledge of tissue-specific mechanisms involved in differential effects on growth factor signaling by fibrillins is scant . In skeletal muscle , a role for fibrillin-1 in regeneration was attributed to the control of TGFβ signaling by fibrillin-1 [16] . Roles for fibrillin-2 in skeletal muscle development and disease have not been studied , especially within the context of growth factor signaling . The investigations presented here address this unexplored area . Results show that loss of fibrillin-2 is accompanied by a decrease in muscle mass and an increase in white fat during the early postnatal period . In vitro and in vivo analyses demonstrate that abnormal activation of BMP signaling contributes to this myopathic phenotype . These results are somewhat surprising , since recent studies have shown that BMP signaling promotes muscle hypertrophy and also protects muscles from denervation-induced wasting [20 , 21] . An additional recent study indicates that loss of BMP signaling promotes intramuscular adipogenesis [22] . In contrast to these recent studies , which rely on the manipulation of cellular components of the BMP signaling pathway , our study focuses on the role of extracellular modulation of BMP signaling . By changing the extracellular environment for BMP signaling , we show that BMP signaling is context dependent . Mutations in extracellular structural macromolecules can cause congenital forms of muscular dystrophy . The congenital muscular dystrophies are a heterogeneous group of disorders that are identified at birth or within the first few years of life . Typical clinical features include hypotonia , delayed motor development , and progressive muscle weakness . The congenital muscular dystrophies caused by merosin ( laminin-2 ) deficiency or defects in collagen VI appear early and are usually severe . Molecular diagnostics over the last two decades have advanced understanding of the muscular dystrophies , but treatment strategies remain limited . Conventional treatments are physical therapy and corticosteroids , and these are aimed at prolonging ambulation and survival . It has been estimated that only half of the congenital muscular dystrophies can be ascribed to known defects [23] . Therefore , it remains important to identify additional muscular dystrophy disease genes , especially if these might lead to novel pathogenetic mechanisms common to at least some of the muscular dystrophies . Common pathogenetic targets might then lead to new opportunities to develop better treatments . Here we present novel data that support the candidacy of FBN2 as a possible congenital muscular dystrophy gene as well as a potential pathogenetic pathway amenable to treatment protocols .
On 129/Sv background , Fbn2 null mice are viable and fertile , but are born with contractures of their forelimbs which disappear within the first week of postnatal life [11] . Because of the similarity between the contractures found in Fbn2 null mice and in humans with CCA , Fbn2 null mice are a model for CCA . Daily inspection of the Fbn2 null forelimbs during the first week of life revealed that the contractures were most severe between birth and P2 and slowly resolved between P3 and P8 ( Fig 1A ) . Since contractures are a common feature of congenital muscular dystrophies , we analyzed Fbn2 null forelimb muscle at birth ( P0 or P1 ) , when contractures were most severe , and at P8 , when contractures had resolved . At P0 , comparable cross sections of forearms demonstrated strikingly reduced areas of skeletal muscle in Fbn2 null mice when compared to wildtype mice ( Fig 1B ) . In order to quantitate the reduction in muscle mass , forearm muscle tissue between the wrist and elbow was dissected and weighed , and the weight of the forearm muscle was normalized to the total body weight . Results showed that the total forearm muscle mass in Fbn2 null mice was reduced by 28% when compared to wildtype ( p = 0 . 001 ) ( Fig 1C , left ) . In addition , when myofibers were counted on comparable sections , the number of myofibers in Fbn2 null muscle areas was reduced by 50% ( p = 0 . 002 ) ( Fig 1C , right ) . H&E staining also showed significant alterations in muscle architecture measured by increased numbers ( 1 . 7 fold , p = 0 . 003 ) of myofibers with centrally located nuclei in Fbn2 null muscle compared to wildtype ( Fig 1D ) . By P8 , when contractures were apparently resolved , H&E staining of Fbn2 null forearm cross sections showed that muscle architecture had improved . The number of myofibers with centrally located nuclei was no longer significantly greater than in wildtype ( Fig 2A , right ) . However , toluidine blue staining on comparable cross sections showed increased fat deposition in the forearms of Fbn2 null mice ( Fig 2B ) . Transmission electron microscopy of fat droplets identified them as white adipocytes—cells containing large lipid droplets with eccentric flattened nuclei ( Fig 2C ) . Skinned and strongly fixed P8 forearms were infiltrated with OsO4 to allow the visualization of both fat and bone as white matter by micro-computed tomography ( μCT ) ( Fig 2D ) ( 16 ) . A series of digital sections demonstrated increased depositions of fat throughout most of the forearm muscle from the wrist to the elbow ( Fig 2D ) . In wildtype forearms , fat was located at the periphery of muscles; in contrast , fat in Fbn2 null forearms was found between muscle and bone and infiltrating muscle bundles ( Fig 2D ) . Forearm muscle was studied in detail , in order to associate muscle abnormalities with contractures . However , whole body scans with μCT were performed using P5 littermates . Hindlimb muscles also showed reduced muscle mass and abnormal fat deposition ( S1 Fig ) . Quantitative real-time PCR ( qPCR ) analysis of P8 forearm muscle showed a significant twofold increase of fat specific markers such as Cebpa , Fabp4 , and Pparg in Fbn2 null muscle compared to heterozygous or wildtype ( Fig 2E ) . These results support the morphological analyses and μCT quantitation of increased fat deposition in the forearm . qPCR analysis of Fbn1 and Fbn2 expression levels showed no abnormal increase in expression of Fbn1 or Fbn2 to compensate for loss of Fbn2 ( S2 Fig ) . Reduced muscle mass , replacement of muscle with fat , centrally localized nuclei in skeletal muscle , and contractures are features associated with muscular dystrophy . Muscle weakness , including grip strength , hindlimb clasping , and poor wire maneuvering , was shown in adult “Mariusz” mice , which harbor an ENU-induced Fbn2 mutation [12] . “Mariusz” mice were generated on a Balb/c X C3H background . Hindlimb clasping and poor hindlimb gait are also characteristic of adult Fbn2 null mice on a 129/Sv background , implying that muscle weakness is a persistent phenotype in this background as well . Since elevated creatine kinase activity is used as a sign of muscle damage , blood samples from Fbn2 null and wildtype mice were tested for creatine kinase activity . Age-matched ( from 2 months to 1 year ) wildtype and Fbn2 null mice were tested . Creatine kinase activity was measured in blood at 6 time points , and the–fold change between wildtype and null mice was averaged . Creatine kinase activity in adult Fbn2 null mice was 2 . 7 fold ( +/- 1 . 4 ) times greater than in wildtype mice ( Table 1 ) . These results are consistent with the persistent muscle weakness observed in adult Fbn2 mice . To evaluate the influence of genetic background , we crossed the Fbn2 null allele from the 129/Sv mouse background onto a C57/Bl6 background by breeding Fbn2+/- mice to wildtype C57/Bl6 mice . After six generations on C57/Bl6 , Fbn2+/- mice were bred together to generate Fbn2 null progeny . Of 47 pups ( 7 litters ) , 12 were null for Fbn2 . Six of the 12 Fbn2 null pups died in the perinatal period . Some of these pups were found alive at birth , but they died soon after birth . Necropsy showed normal lungs and heart with no blood in the chest cavity . Central cyanosis suggested that the Fbn2 null pups died from hypoventilation . Gross observation of the diaphragm at P0 , which looked thin and frail , indicated reduced muscle mass , and H&E staining showed poor muscle architecture ( Fig 3A ) . Two Fbn2 null mice lived to P25 and P40 , when they were sacrificed and analyzed by μCT . These mice were clearly smaller than their wildtype littermates ( Fig 3B and 3C ) . μCT showed that accumulation of fat was not present in the subcutaneous tissues ( Fig 3B ) but was rather limited to the forelimb connective tissue . Hence , poor muscle architecture and the specific replacement of muscle tissue with fat were similar in Fbn2 null mice on 129/Sv and C57/Bl6 backgrounds , but the myopathy was less severe on the 129/Sv background . These results indicate the importance of genetic modifiers on the severity of Fbn2 myopathy . Results obtained after 6 generations on C57/Bl6 were not uniform . In addition , the possibility remained that further backcrossing into the pure C57/Bl6 background might lead to fetal lethality . After 9 generations on C57/Bl6 , 4 litters with 21 total pups contained 4 Fbn2 null pups , each of which died at P0 . These results indicate that , when the Fbn2 null allele is on a pure C57/Bl6 background , death of Fbn2 null mice will occur shortly after birth . To assess possible effects on fibrillin-1 , forearm muscles from Fbn2 null and wildtype mice were dissected and digested with pepsin , following an established protocol [24] , and pepsin-resistant fibrillin-1 fragments were analyzed over the first 8 days of postnatal life . We identified a pepsin-resistant fragment of fibrillin-1 with a Mr of 70 kDa ( Fig 4A ) similar to previously characterized fibrillin-1 fragments from human amnion [24] . At P0 , when contractures were severe and muscle mass was reduced , no pepsin-resistant fibrillin-1 fragments were detected . However , from P1 to P8 , fibrillin-1 bands ( asterisk ) of equal intensities were detected in wildtype , Fbn2+/- , Fbn2-/- muscle , indicating that the amount of fibrillin-1 in forearm muscle tissue was not significantly altered by Fbn2 genotype ( consistent with qPCR data shown in S2 Fig ) . Since no pepsin-resistant fibrillin-1 fragments were identified in P0 tissue digests , fibrillin-1 microfibrils may undergo maturation processes after birth that are necessary to stabilize certain fibrillin-1 peptides to pepsin digestion . As a positive control for the pepsin digestions , Ponceau staining showed two bands ( open stars ) , most likely the α1 and α2 chains of type I collagen , that were present in all extracts from P0 to P8 . The Ponceau staining also revealed a pepsin-resistant band around 40 kDa ( Fig 4A ) . Surprisingly , the intensity of this band correlated with the Fbn2 genotype when contractures were severe ( P1-P3 ) . Quantitation of this band in P1 and P3 Fbn2 mutants compared to wildtype was performed , with normalization relative to the pepsin-resistant collagen bands . These data showed a significant reduction in both heterozygous and null forearms compared to wildtype ( S3 Fig ) . N-terminal sequence analysis identified this band as myosin heavy chain polypeptide 8 ( Myh8 ) , a perinatal myosin present in skeletal muscle . To test whether loss of fibrillin-2 leads to reduced expression of Myh8 , RNAi mediated knockdown of Fbn2 was performed in C2C12 cells during myoblast to myotube differentiation . After 2 days , knockdown of Fbn2 resulted in downregulation of Myh8 while other myogenic factors like Myogenin and MyoD were not affected ( Fig 4B ) . Similarly , knockdown of Fbn2 had no effect on the expression of Fbn1 ( S4 Fig ) . Taking all the data together , we conclude that severe contractures in Fbn2 null mice are associated with reduced skeletal muscle mass , abnormal muscle histology , and a delay in expression of the perinatal Myh8 . Interestingly , mutations in MYH8 cause the Trismus-pseudocamptodactyly syndrome ( a Carney complex variant ) in humans , characterized by contractures in hands and feet [25] . Therefore , contractures in the Fbn2 null mice may be caused by a delay in muscle differentiation . BMP signaling is thought to cause a delay in muscle differentiation [26] . Hence , we tested whether increased BMP activity has an effect on Myh8 expression . When C2C12 cells were utilized , Myh8 expression was 3- to 10-fold downregulated after 24 hours of treatment with BMP-4 or BMP-7 ( Fig 4C ) . This is consistent with previous observations that BMP treatment of C2C12 cells inhibited the formation of multinucleated myotubes and resulted in downregulation of myosin heavy chain expression [26] . In addition , these results suggest that BMP treatment of C2C12 cells has a similar negative effect on Myh8 expression as RNAi-mediated knock down of Fbn2 during myoblast to myotube formation . We therefore hypothesized that the increased numbers of muscle cells with centrally located nuclei in Fbn2 null forearms were due to a delay in differentiation caused by activation of BMP signaling . To test this hypothesis , we analyzed cross sections of Fbn2 null and wildtype forearms by immunofluorescence using antibodies against phospho-Smad ( pSmad ) 1/5/8 and pSmad 2/3 , respective downstream mediators of BMP and TGFβ signaling . At early postnatal time points when contractures are severe , we detected abundant signals for pSmad 1/5/8 in Fbn2 null , but not in wildtype , muscles ( Fig 5A , upper panel ) . At P8 , when contractures are resolved , pSmad 1/5/8 signals in Fbn2 null forearm muscles were no longer detectable . In support of positive pSmad 1/5/8 staining , qPCR results for Id1 , a BMP-responsive gene , showed a two-fold increase in P0 Fbn2 null forearm extracts compared to wildtype ( S5 Fig ) . To exclude that activation of BMP signaling was due to upregulation of BMP expression , we measured levels of BMP mRNAs in Fbn2+/- , Fbn2-/- , and wildtype forearms by qPCR , but could not detect any significant differences in the mutants , when compared to wildtype ( Fig 5B ) . We also used qPCR and external template standards of known concentrations for each Bmp qPCR primer set , in order to determine which of the Bmps are most highly expressed in skeletal muscle . Bmp4 and 7 had the highest expression in mouse forearm muscles at P0 ( Fig 5C ) . To test the hypothesis that activated BMP signaling causes the muscle phenotype observed in Fbn2 null forelimbs , we treated null and wildtype forelimbs in organ culture with Noggin , a BMP inhibitor . Exogenous Noggin was titrated ( 200–800 ng/ml ) , and the number of myofibers with central nuclei was counted . Multiple fields from different H&E stained sections were counted , and the numbers were graphed as percentage of the total number of myofibers . Noggin treatment at 6 nM ( 200 ng/ml ) for 2 days in organ culture was sufficient to significantly improve the muscle architecture , measured as the decrease of myofibers with central nuclei from 42 to 14% ( p = 0 , 004; Fig 6A ) . In contrast , when control Fbn2+/- forelimbs were either treated with 6 nM ( 600 ng/ml ) exogenous BMP-7 complex or were untreated for 2 days in organ culture , a significant increase of myofibers with central nuclei ( from 20% in the untreated forelimb to almost 60% in the treated forelimb , p<0 . 001 ) was observed ( Fig 6B ) . Since Bmp7 was highly expressed in P0 mouse forelimb muscle and since addition of BMP-7 to forelimb organ cultures resulted in an increase in the numbers of myofibers with centrally localized nuclei , we hypothesized that lowering the expression level of Bmp7 should have beneficial effects on the Fbn2 null myopathy . To test this hypothesis , Fbn2+/- mice were crossed with Bmp7+/- mice [27] to generate Fbn2+/-; Bmp7+/- double heterozygous mice . Fbn2+/-; Bmp7+/- mice were then crossed with each other to generate Fbn2-/-; Bmp7+/- mice . Analysis and quantitation of forelimb muscle and fat was performed by μCT , after fixation and incubation of limbs with OsO4 . A series of digital sections from Fbn2+/-; Bmp7+/- forelimb is shown in Fig 7A . The percentages of fat on comparable μCT sections of Fbn2 null and wildtype littermates were quantitated relative to the areas of bone and muscle . Fat in Fbn2 null forearms was significantly increased by two-fold over wildtype ( p = 0 . 003; Fig 7B ) . Analysis of Fbn2-/-; Bmp7+/- forelimbs showed that the amount of fat and muscle on serial digital cross sections returned to normal wildtype levels ( Fig 7B ) . This genetic approach provided further evidence that activated BMP signaling caused the Fbn2 null myopathy , including both the reduction in muscle mass and the increase in fat that infiltrates the forelimb muscle . It was previously shown that loss of fibrillin-2 results in syndactyly , likely due to a loss of BMP activity in the interdigital space [11] . In the developing interdigital space , loss of fibrillin-2 is accompanied by a loss of fibrillin-1 ( S6 Fig ) . However , in early postnatal forelimb muscle , both fibrillins are detectable by immunostaining , and loss of fibrillin-2 is not accompanied by any substantial loss or redistribution of fibrillin-1 ( S7 Fig ) . Moreover , loss of fibrillin-2 in the forelimb is accompanied by activation of BMP signaling ( Fig 5A ) . Therefore , control of BMP activity by fibrillin cannot be explained by the simple presence or absence of fibrillin . To gain further insight into the molecular mechanisms by which fibrillin-2 might control BMP activity , we performed in vitro experiments . First , we measured the mRNA expression levels of Fbn1 and Fbn2 during mouse C2C12 myoblast to myotube differentiation . Fbn2 mRNA was 15-fold increased 2 days after initiation of differentiation when compared to levels at the initiation start , while Fbn1 mRNA expression was increased only 2 . 5-fold ( Fig 8A ) . In a second experiment , we performed the C2C12 myoblast to myotube differentiation assay while Fbn2 expression was knocked down by RNAi . Two days after initiation of differentiation the cell culture supernatant was removed and added to C2C12BRA cells , which harbor the BMP-inducible plasmid BRE-luc and serve as reporter cells for BMP bioactivity [28] . Supernatant from cells subjected to Fbn2 RNAi contained 4 times more BMP bioactivity than supernatant from control cells ( Fig 8B ) . These results were consistent with in vivo observations of increased BMP activity in Fbn2 null forelimb muscle . In addition , the in vitro results indicate that , in the process of muscle maturation , fibrillin-2 performs a much more predominant role than fibrillin-1 . Based on the data , it seemed likely that fibrillin-2 is a negative regulator of BMP activity in muscle . We therefore hypothesized that binding to fibrillin-2 might confer latency to the BMP complex . To test this hypothesis , we created an assay using different methods for presenting BMP-7 complex to C2C12 cells . The positive control condition was addition of BMP-7 complex to the medium ( BMP-7 complex in solution ) . To immobilize BMP-7 , we tested 3 different substrates adsorbed onto plastic wells: ( 1 ) rF86 , an N-terminal recombinant polypeptide of fibrillin-2 ( Fig 8C ) ; ( 2 ) an antibody ( mAb 689 ) coupled to rF86 ( Fig 8D ) ; ( 3 ) a capture antibody against the N-terminal His6-tag of the BMP-7 prodomain ( Fig 8D ) . The rF86 polypeptide was chosen because it displayed high affinity binding ( Kd = 8–25 nM ) in previous interaction studies with BMP prodomains [6 , 29] . We adsorbed mAb689 , which we previously described as a pan-fibrillin antibody that recognizes a site close to EGF4 [9] , to the well and then coupled rF86 to mab689 . We then immobilized BMP-7 complex to the adsorbed substrates . To quantitate and compare the efficiency of BMP-7 immobilization through the two different capture approaches , the immobilized BMP-7 complex was stripped and compared to standards of known concentrations of BMP-7 complex in a dot blot immunoassay . mAb2 , specific for the BMP-7 prodomain [5] was used for detection . Both capture approaches resulted in immobilization of quantitative amounts of BMP-7 complex , although rF86 ( coupled to mAb689 ) captured 3-fold more BMP-7 complex than the anti-His6-tag antibody . In order to measure BMP bioactivity , C2C12 cells were seeded onto the immobilized BMP-7 complexes captured by the two different approaches . After 6 hours , the total mRNA was immediately harvested , and mRNA expression levels of Id3 , a BMP responsive element , were monitored by qPCR . BMP-7 complex simply added to the medium and BSA coated wells ( “untreated” ) served as positive and negative controls . Immobilizing 1 . 7 ng/well of BMP-7 complex , using the His6-tag , resulted in 18-fold induction of Id3 , which was equivalent to adding 30 ng/well BMP-7 complex to the medium ( Fig 8E ) . This demonstrated a more than 15-fold increase of BMP-7 activity when BMP-7 is localized to a scaffold at the bottom of the well compared to BMP-7 in solution . However , immobilizing BMP-7 complex through rF86 bound to the dish reduced BMP bioactivity to negative control levels . Binding of BMP-7 complex to fibrillin-2 followed by stripping off BMP-7 did not result in any reduction of BMP-7 bioactivity ( S8 Fig ) , indicating that BMP-7 is reversibly inactivated by binding to fibrillin-2 .
Collectively , our investigations demonstrate that the absence of fibrillin-2 causes a delay in forelimb muscle differentiation , contractures that resolve as muscle architecture improves , and infiltration of fat into the limb connective tissue space . Neonatal contractures in fibrillin-2 null mice were associated with a delay in the appearance of a specific perinatal myosin heavy chain . Mutations in the gene ( MYH8 ) encoding this perinatal myosin heavy chain cause a human syndrome characterized by distal contractures , indicating that contractures in fibrillin-2 null mice may be the result of effects on myosin heavy chain-8 . In vitro and in vivo experiments showed that abnormal activation of BMP signaling was the likely reason for the delay in muscle differentiation , including the delay in expression of myosin heavy chain -8 , and for fatty infiltration of the limb connective tissue . The finding that activation of BMP signaling caused decreased muscle mass in Fbn2 null mice is surprising , since genetic ablation of Smad4 in adult mice caused muscle atrophy and weakness in response to denervation , while overexpression of an activated form of Alk3 ( a type I BMP receptor ) could rescue the Smad4 muscle phenotype and overexpression of noggin ( an inhibitor of BMP signaling ) caused muscle hypertrophy [20] . Similarly , when BMP-7 was overexpressed in muscle , muscle mass was increased along with muscle hypertrophy [21] . The major difference between these studies and ours is that , in our study , activated BMP signaling is the result of a deficient environment and not the result of overexpression of a BMP ligand or inhibitor . Furthermore , our study focuses on very early postnatal muscle development , when fibrillin-2 performs an important role , whereas these other studies [20 , 21] examine the role of BMP signaling in adult mice , particularly in response to denervation . When Bmpr1a ( Alk3 ) was knocked out in Myf5 Cre expressing cells , mice were born runted and muscle mass was reduced , but muscles appeared unaffected when MyoD Cre was used to knockout Bmpr1a [22] . These latter results are consistent with the concept that BMP signaling may exert different effects in different microenvironments . In contrast to these studies that directly manipulate known components of the BMP signaling pathway , our study shows the exquisite sensitivity of BMP signaling to the fibrillin-2 extracellular environment during early postnatal muscle development . TGFβ signaling is known to be context-dependent . TGFβ has been described as a “cellular switch” that provides “a mechanism for coupling a cell to its environment” [30] . Results presented here indicate that fibrillin-2 is an important constituent of the skeletal muscle environment and that BMP signaling is coupled to the fibrillin-2 microenvironment . Previously , syndactyly in Fbn2 null mice was associated genetically with a loss of BMP signaling [11] . In the studies presented here , loss of fibrillin-2 in the interdigital space was accompanied by a loss of fibrillin-1 ( S6 Fig ) , but in P0 skeletal muscle , fibrillin-1 seemed grossly similar to wildtype in the absence of fibrillin-2 ( S7 Fig ) . More information is required in order to know why loss of fibrillin-2 in the early postnatal skeletal muscle environment leads to activation of BMP signaling . As a first step , we addressed whether direct binding to fibrillin-2 can regulate BMP signaling . In previous studies , we demonstrated that BMP growth factors form complexes with their prodomains and that the prodomain targets the prodomain/growth factor complex to specific sites on fibrillins [5 , 6 , 29] . We also showed that some BMP complexes ( BMP-4 , -5 , and -7 ) are as bioactive in solution as the free BMP growth factor dimer [6 , 29] , suggesting that unlike the prodomains of TGFβ , GDF8 , BMP-10 , and other members of the TGFβ superfamily [29 , 31–33] , BMP prodomains do not perform an intrinsic inhibitory role [34] . Recently , the crystal structure of BMP-9 prodomain/growth factor complex provided a molecular description of active BMP complexes such as BMP-9 and BMP-7: these complexes share an “open” conformation” [35] , unlike the TGFβ small latent complex , which is constrained by the propeptides into a more closed conformation [36] . Here in this study , we found that , when BMP-7 complex is targeted to a solid substrate ( by mAb recognizing the His6-tag on the prodomain ) close to cells , it is potentiated . We interpret these results to suggest that spatial concentration of BMPs to a protein scaffold close to cells dramatically increases their bioactivity when compared to the same numbers of freely diffusing BMP molecules . In addition , we showed that , when BMP-7 complexes are bound to a fibrillin-2 scaffold , BMP bioactivity was inhibited . This inhibition by fibrillin binding is most likely induced by a conformational change in the prodomain structure that prevents access of BMP receptors to the growth factor . Interestingly , molecular modeling indicated that BMP-9 prodomain/growth factor complex can adopt a closed conformation similar to the TGFβ small latent complex [35] . Our ongoing studies are addressing whether binding to fibrillin can induce such a conformational change in BMP-7 complex . Our new mechanistic insights into the control of BMP bioactivity by fibrillins have important implications: if BMP complexes are latent when bound to fibrillin , then there must be mechanisms for activating BMP signaling in appropriate cellular contexts . These mechanisms are unknown , since it has been thought that BMPs do not require activation . It is possible that cells may activate fibrillin-bound BMP complexes through mechanical means , either pulling on the prodomain , as in the case of integrin pulling to activate latent TGFβ [36] or pulling on fibrillin itself to locally release BMP complexes close to cellular BMP receptors . This activation mechanism may be differentially utilized by cells , depending on their expression of appropriate cellular receptors . We currently do not know what cellular receptors may activate BMP complexes bound to fibrillin . Although we cannot currently explain why BMP signaling is activated in the skeletal muscle of fibrillin-2 null mice , we can speculate . One possible explanation is that , in the absence of fibrillin-2 , BMP complexes are not bound by fibrillin-2 and therefore not inhibited . However , our in vitro experiment suggests that free , diffusing BMP complexes would be much less potent than the same quantities of targeted BMPs . A second possibility is that , in the absence of fibrillin-2 , BMP complexes are mistargeted to fibrillin-1 . On fibrillin-1 , without fibrillin-2 present , BMP complexes may be more concentrated than usual and potentially more potent . The conformation of fibrillin-1 microfibrils , in the absence of fibrillin-1 , may lead to easier activation of BMPs , particularly during the early postnatal period when we were able to detect differences in resistance to pepsin . Further maturation and stabilization of fibrillin-1 microfibril conformation during the first week of postnatal life would be consistent with this explanation , since activation of BMP signaling is limited to the first few days after birth . In addition , fibrillin-1 may occupy specific sites , vis a vis certain muscle cells , distinct from those occupied by fibrillin-2 . Future investigations will be directed toward elucidating the spatially specific muscle cell niches defined by fibrillin-1 compared to fibrillin-2 . Our results can be taken in several translational directions . First , the myopathy caused by the absence of fibrillin-2 in mice may be relevant to human congenital muscular dystrophies . In contrast to heterozygous mutations in Fbn2 causing CCA , with contractures and muscle weakness , results in mice predict that homozygous Fbn2 mutations may cause a form of human congenital muscular dystrophy . In humans , an autosomal recessive form of congenital muscular dystrophy with arthrogryposis ( OMIM %253900 ) has been reported . The most recent report [37] described a family in which a child died of respiratory failure two hours after birth . Although the lungs were normal , the skeletal muscle was replaced by adipose and some connective tissue . Skeletal muscle from another fetus , electively terminated when no movements were detected after 19 weeks of gestation , showed multiple contractures and severe skeletal muscle wasting with adipose tissue replacement . Fbn2 null mice on the C57Bl/6 background may model this type of congenital muscular dystrophy: death just after birth , contractures , skeletal muscle wasting with adipose tissue replacement , normal lungs , and autosomal recessive inheritance . A second translational direction is the potential for identification of genetic modifiers of congenital muscular dystrophy . LTBP4 has been identified as an important genetic modifier of muscular dystrophy in mice [38] and in boys with DMD [39] . It is interesting to note here that LTBP-4 is a fibrillin-like molecule that is incorporated into fibrillin microfibrils . Similar approaches could be used to identify the genetic modifier ( s ) in 129 vs . C57/Bl6 mice controlling severity of the Fbn2 myopathy . These genetic modifiers could then be tested to determine utility in the clinic to predict severity of common forms of muscular dystrophy . Third , since fibrillin microfibrils integrate basement membrane molecules with the adjacent connective tissue , it is possible that defects in cell-matrix interactions that underlie congenital muscular dystrophies may perturb fibrillin microfibrils and cause dysregulation of BMP signaling . Adenoviral overexpression of Noggin in the muscle of mdx mice ( a model of DMD ) resulted in a reduction in fibrotic/necrotic area and an improvement in muscle histology [40] , indicating that abnormally activated BMP signaling may contribute to DMD and perhaps to other congenital forms of muscular dystrophy . However , mechanisms underlying this improvement of mdx muscle were not investigated . Our new results provide a potential mechanism by which BMP signaling may be activated in some forms of congenital muscular dystrophy , if perturbation of fibrillin microfibrils is a general or specific consequence of defects in muscle cell-matrix interactions . Fourth , a better understanding of the molecular mechanisms by which fibrillins control BMP signaling is required before therapeutics can be designed to specifically target the muscle cell microenvironment . We have shown that the absence of fibrillin-2 has tissue-specific effects on the skeletal muscle , but not on other tissues . If activated BMP signaling contributes to myopathy and infiltrating fat , then modulating the fibrillin-2/BMP interaction may be a clever therapeutic strategy . Lastly , our results advance knowledge regarding potential cellular mechanisms by which muscle is replaced with fat , a key clinical feature of muscular dystrophy . Here we showed that active BMP-7 signaling performs a critical role in the formation of white fat . BMP-7 is known to play a role in the induction of brown adipocytes from progenitor cells residing within skeletal muscle [41 , 42] . In contrast to these primarily in vitro studies , our results demonstrated that ablation of one allele of Bmp7 is sufficient to rescue the induction of white fat in Fbn2 null skeletal muscle , demonstrating an in vivo role for BMP-7 in the transdifferentiation of cells from muscle to white fat . Whether this role is specific for BMP-7 is currently under investigation . Genetic approaches will be used in the future to reveal the identity of the BMP-7 responsive cells that are induced to become white adipocytes .
Fbn2 null mice [11] were maintained on a 129/Sv background and genotyping by PCR was performed as previously described [3] . Bmp7+/- mice were maintained on a 129S/SvEv-Gpi1c background . For genotyping , wildtype and mutant Bmp7 alleles were amplified by PCR as previously described [27] . Fbn-/- mice were bred to Bmp7+/- mice to yield Fbn2+/-; Bmp7+/- mice . Doubly heterozygous mutant mice were bred to each other to generate Fbn2-/-; Bmp7+/- , and all progeny were genotyped by PCR . Fbn2+/- mice on 129/Sv background were bred to wildtype C57/Bl6 mice , and Fbn2+/- progeny were bred to wildtype C57/Bl6 mice for nine generations . Fbn2+/- mice were bred to each other to generate Fbn2-/- mice after six and nine generations of backcrossing . Maintenance of mice and all experimental procedures were conducted in accordance with National Institutes of Health guidelines and were approved by the OHSU IACUC ( approval number IS00003821 ) . Serial 4 μm cryosections of frozen muscle were prepared and stained with hematoxylin and eosin ( H&E ) using the Shandon Rapid Chrome H&E frozen section staining kit ( Thermo Scientific , Waltham , MA ) according to the manufacturer’s protocol . Sections were examined using a Zeiss Axiophot microscope , and micrographs were recorded digitally using AxioVision software ( version 4 . 5; Zeiss , Germany ) . In addition , fresh tissues were fixed in 1 . 5% glutaraldehyde/1 . 5% paraformaldehyde with 0 . 05% tannic acid in cacodylate buffer , followed by 1% buffered OsO4 and then rinsed , dehydrated , and embedded in Spurr’s epoxy . Thick sections ( 8 μm ) from these strong-fixed tissue blocs were stained with Toluidine blue O stain ( Sigma , St . Louis , MO ) and examined by light and electron microscopy . Immunofluorescence microscopy was performed as previously described [3 , 43] . Primary antibodies , polyclonal rabbit anti-phospho Smad1/5/8 ( 9511 , Cell Signaling , Danvers , MA ) or anti-phospho Smad2/3 ( 3101 , Cell Signaling ) , and secondary antibodies , Alexa 488 goat anti-rabbit or Alexa 568 goat anti mouse ( Invitrogen Molecular Probes , Eugene , OR ) , were used at a concentration of 1:1000 . Citrated plasma was stored at -80°C until used . Plasma was assayed using a colorimetric Creatine Kinase Activity Assay Kit , according to the manufacturer’s protocol ( Abcam , Cambridge , MA ) . Briefly , in a 96-well plate an NADH standard curve was developed ( 0–10nmol ) , along with blood samples . 5 μl of sample were brought to 50 μl with addition of assay buffer . 50 μl of a reaction mix containing assay buffer , enzyme mix , developer , ATP , and substrate was added to each well . The plate was incubated at 37°C for 40 minutes with spectrophotometric readings at 450 nm taken every 2 minutes . Each sample was tested in one well of a 96-well plate . Multiple dilutions of each sample were tested to ensure that readings were within the standard curve range . Creatine kinase activity was calculated by the equation: ( nmol NADH/ ( reaction time X sample volume in ml ) ) X dilution factor = nmol/min/ml . Pepsin digests of mouse P0-P8 forearm muscle was performed as previously described [24] . In brief , total forearm muscle was dissected , dropped in liquid nitrogen , and ground with a mortar and pestle . The pulverized tissue was first washed with cold water , then three times with cold 1 M NaCl ( 20 μl per mg tissue ) , and once more with water . The final pellet was resuspended in 0 . 5 M acetic acid and the stirred suspension was digested with pepsin ( Sigma; 8 μg of pepsin/1 mg wet weight ) for 16 h at 4°C . The solubilized material was collected following centrifugation . Sodium chloride ( 10% , w/v ) was added to the supernatant , and the precipitate was collected by centrifugation . The pellet was resuspended in 0 . 1 M Tris-HCI , pH 8 . 1 , and stirred at 4°C for at least 72 hours to inactivate the pepsin . Equal protein amounts extracted from Fbn2 +/+ , Fbn2+/- , and Fbn2-/- muscle ( determined by using the Pierce BCA protein assay kit ( Pierce , Rockford , IL ) ) were loaded onto a 4–20% polyacrylamide gel and analyzed under non-reducing conditions , followed by Western blotting using pAb 9543 , specific for fibrillin-1 . For quantitation of the Myh8 protein band , Ponceau S stained membranes were scanned , and band intensities were measured using NIH ImageJ ( Rasband ) ; results were normalized relative to the pepsin resistant collagen band . Results shown were generated from N = 3 litters containing littermates of each genotype for P1 , and N = 2 litters for later time points . Mouse C2C12 cells ( CRL-1772 , ATCC , Manasassas , VA ) were cultured in Dulbecco’s modified Eagle’s medium ( DMEM ) ( MediaTech , Hendon , VA ) supplemented with 10% fetal bovine serum ( Atlanta Biologicals , Lawrenceville , GA ) and penicillin/streptomycin ( MediaTech ) . Cells were seeded on Costar 6 well plates ( Sigma ) at 50 , 000 cells/well and transfected at 30% and 60% confluency with 100 μM Fbn2 siRNA and scrambled control using Lipofectamine reagent according to the manufacturer’s instructions ( all reagents from Invitrogen ) . At 90% confluency , cells were stimulated with DMEM containing 2% horse serum for 48 hours . The cell culture supernatant was analyzed for BMP activity by pipetting it onto stably transfected cells carrying BMP-responsive elements from the Id1 promoter fused to a luciferase reporter gene , as previously described [28] . Total RNA of the cell layer was harvested using 1 ml TRIzol reagent ( Invitrogen ) , and expression of myogenic markers was monitored by real-time quantitative PCR . 1–10 mg of dissected P0 or P8 mouse forearm muscle was dropped into 0 . 8 ml of TRIzol reagent ( Invitrogen , Carlsbad , CA ) and ground with a micropestle ( Kimble-Chase , Vineland , NJ ) into smaller pieces . RNA extraction was performed according to the manufacturer’s protocol . A subsequent sample purification step was included using the RNeasy kit ( Qiagen , Valencia , CA ) , and residual DNA contamination was removed from each sample by using the Turbo DNA-free kit ( Ambion , Austin , TX ) . RNA samples were quantified by photospectrometry , and 0 . 1 to 1 . 0 μg of RNA per sample was reverse transcribed using the Bio-Rad iScript cDNA synthesis kit ( Bio-Rad , Hercules , CA ) . Samples were amplified in triplicates using the iTaq SYBR Green Supermix ( Bio-Rad ) in an iQ5 Multicolor Real-Time PCR Detection System ( Bio-Rad ) . Analysis of data was performed using the 2-ΔΔCt method [44] and quantitated relative to the ARBP0 gene . Gene expression was normalized to littermate wildtype control mice , which provided an arbitrary constant for comparative fold expression . Primers used for qRT-PCR are listed in Table 2 . For quantitation of absolute BMP mRNA expression levels , 500 bp PCR fragments were generated which served as external standard templates for each of the BMP qPCR primer sets utilized . Real-time PCR reactions using a dilution series of known concentrations for each individual standard were simultaneously run next to reactions using reverse transcribed mRNAs from P0 mouse forearm muscles as templates . Standard curves were generated by graphing threshold cycles of standard PCR reactions against the concentrations of the standard templates . The absolute amounts of each BMP mRNA were then calculated by comparing threshold cycles obtained from reactions using reverse transcribed forearm RNAs as templates with the standard curves . Sequences of primers used to generate the 500 bp standard templates are listed in Table 3 . Limbs from neonatal mice were skinned , trimmed proximally and distally , washed 2 times in DMEM containing penicillin/streptomycin , and cultured for 48hs in DMEM with 10% fetal bovine serum containing either 200–800 ng mouse Noggin ( R&D Systems , Minneapolis , MN ) or 600–2400 ng BMP-7 complex [5] on a rocking platform . Untreated samples served as controls . Three independent organ culture experiments were carried out and analyzed . Each time , forelimbs from different animals were used . All animals were littermates . μCT analysis of epoxy embedded limbs was performed using a Scanco μCT 35 instrument ( Scanco Medical , Basserdorf , Switzerland ) , according to the manufacturer’s instructions . Skinned forelimbs were fixed and treated with OsO4 as described [45] . Series of digital limb cross sections were generated by μCT . Areas of fat , muscle and bone were quantitated using NIH ImageJ ( Rasband ) and computed as percentages of the total area of the cross section . mAb 689 ( 17 . 4 μg/ml in coating buffer ) was adsorbed overnight onto all wells of an ELISA plate . 5% milk was incubated as a blocking solution for 1 hour . Then , the wells were washed three times with TBS/Tween and subsequently incubated with 0 . 5 μM rF86 in 2% milk for 2 hours . The plate was washed again and incubated with 47 . 6 μg/ml BMP-7 complex ( 0 . 5 μM ) in 2% milk for 2 hours . After final washes , the plate was incubated with 100 μl/well of 0 . 1M Glycine , pH 2 . 3 , for 20 minutes in order to strip off the bound BMP-7 complex . The contents of all 96 wells were pooled and dialyzed against 0 . 1M acetic acid overnight . This solution was lyophilized and resuspended in 50 μl TBS . 5 μl dots were placed on a nitrocellulose membrane together with a diluted series of dots containing BMP-7 complex at known concentrations ( the standard curve ) . After drying , the membrane was blocked in 5% milk and incubated with anti-BMP-7 prodomain antibody mAb2 ( 1 mg/ml , 1:500 ) for 2 hours . The membrane was washed and subsequently incubated with an HRP-conjugated goat anti-mouse antibody for 2 hours . After the final washes , signals were developed using the Bio-Rad Opti 4CN Substrate kit . The membrane was scanned and signals were quantitated using image quant software . A similar procedure was performed after first adsorbing an anti-His6-tag mAb onto all wells of an ELISA plate . C2C12 cells were seeded onto the plates in which quantitated amounts of BMP-7 complex had been immobilized . After 6 hours , the total mRNA was immediately harvested , and mRNA expression levels of Id3 were monitored by qPCR . BMP-7 complex was added to the medium to serve as a positive control , and BSA coated wells served as negative controls . Statistical analysis of muscle mass measurements , myofibril counts on H&E stained sections , and muscle and fat quantitations in P8 muscles were generated with GraphPad Prism 5 . 0 for Windows ( GraphPad , San Diego , CA ) . P values were obtained from a One-way Analysis of Variance ( 1-way ANOVA ) with Significance level Alpha = 0 . 05 . | New strategies for treating congenital muscular dystrophies are needed . Current treatments are limited and aim to prolong ambulation and survival . Since most of the genes responsible for congenital muscular dystrophies are still unknown , elucidation of these genes may provide new insights that can lead to novel treatments . Fibrillin-2 null mice are born with myopathy and contractures and demonstrate accumulation of white fat during the early postnatal period . Both the histological features of myopathy and the accumulation of fat are rescued by inhibiting BMP signaling . Results indicate that FBN2 is a candidate gene for congenital muscular dystrophy and that strategies aimed at inhibition of abnormal BMP signaling may be applicable to muscular dystrophies . Furthermore , results reveal the importance of extracellular control of BMP signaling in skeletal muscle . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2015 | Abnormal Activation of BMP Signaling Causes Myopathy in Fbn2 Null Mice |
Plants actively perceive and respond to perturbations in their cell walls which arise during growth , biotic and abiotic stresses . However , few components involved in plant cell wall integrity sensing have been described to date . Using a reverse-genetic approach , we identified the Arabidopsis thaliana leucine-rich repeat receptor kinase MIK2 as an important regulator of cell wall damage responses triggered upon cellulose biosynthesis inhibition . Indeed , loss-of-function mik2 alleles are strongly affected in immune marker gene expression , jasmonic acid production and lignin deposition . MIK2 has both overlapping and distinct functions with THE1 , a malectin-like receptor kinase previously proposed as cell wall integrity sensor . In addition , mik2 mutant plants exhibit enhanced leftward root skewing when grown on vertical plates . Notably , natural variation in MIK2 ( also named LRR-KISS ) has been correlated recently to mild salt stress tolerance , which we could confirm using our insertional alleles . Strikingly , both the increased root skewing and salt stress sensitivity phenotypes observed in the mik2 mutant are dependent on THE1 . Finally , we found that MIK2 is required for resistance to the fungal root pathogen Fusarium oxysporum . Together , our data identify MIK2 as a novel component in cell wall integrity sensing and suggest that MIK2 is a nexus linking cell wall integrity sensing to growth and environmental cues .
Plant cells are surrounded by a thick cell wall that is composed primarily of complex carbohydrates [1] . The cell wall plays a pivotal role in plants , as it provides the mechanical strength that allows the plant to resist both external and internal ( turgor ) pressure , protects the cell from biotic and abiotic stresses , and forms the interface between neighbouring cells [1] . The main load-bearing elements of the cell wall are cellulose microfibrils , which are interconnected with a matrix consisting of hemicelluloses , pectins , and a small amount of structural proteins [1] . To allow cell expansion and growth as well as to provide protection against biotic and abiotic stress , the plant requires the ability to adjust the chemical and mechanical properties of the cell wall , for which it requires feedback information about wall integrity . Yeast cells possess an active cell wall integrity ( CWI ) maintenance mechanism that monitors the status of the cell wall and activates compensatory responses upon damage [2] . Evidence is emerging that plants also have an active CWI sensing mechanism [1 , 3–8] . In plants , cell wall damage can be induced in a controlled manner through pharmacological or genetic inhibition of the cellulose synthase complex [1 , 3 , 5] . Disruption of CWI through inhibition of cellulose biosynthesis results in activation of several stress responses including production of reactive oxygen species [9] , jasmonic acid ( JA ) , salicylic acid ( SA ) , and ethylene [10 , 11] , changes in cell wall composition including lignin deposition [12 , 13] , callose deposition [13] , and alterations in pectin methyl-esterification status [14–16] , and finally swollen roots and growth inhibition [17] . Interestingly , these stress responses are reminiscent of the plant’s defence reaction to pathogens and insects [1 , 3 , 5 , 6 , 18] . The initiation of the plant’s defence response against pathogens requires perception of pathogen-associated molecular patterns or damage-associated molecular patterns through plasma membrane-localized receptor kinase ( RK ) proteins [19] . These RK proteins contain an extracellular ligand binding domain , a single-pass transmembrane domain , and an intra-cellular kinase domain [20] . Analogous to their role in pathogen recognition , RKs could be ideal candidates as sensors of CWI , as they allow signal transmission from the external environment to the inside of the cell . In the model plant Arabidopsis thaliana ( At , hereafter referred to as Arabidopsis ) , the family of RKs contains over 400 members [21] . Several RKs have been identified as putative CWI sensors [1 , 4–8 , 22] , among them the cell surface-localized RK THESEUS1 ( THE1 ) [23] . THE1 was identified in a screen for suppressors of prc1-1 , a mutant in the cellulose synthase subunit CesA6 [23] , and belongs to the malectin-like Catharanthus roseus Receptor-Like Kinase 1-like ( CrRLK1L ) family [4] . While the cellulose-deficient mutant prc1-1 displays constitutive growth inhibition and lignin deposition , these phenotypes were partially relieved in the prc1-1 the1-1 double mutant [23] . As the1-1 does not impact cellulose biosynthesis in prc1-1 mutant background , it was suggested that THE1 functions as a CWI sensor [23] . The CrRLK1L family contains 17 members in Arabidopsis , and besides THE1 , includes FERONIA/SIRENE ( FER/SRN ) , HERCULES1 ( HERK1 ) , HERCULES2 ( HERK2 ) , ANXUR1 ( ANX1 ) , ANXUR2 ( ANX2 ) , ERULUS/[CA2+]CYT-ASSOCIATED PROTEIN KINASE 1 ( ERU/ CAP1 ) and CURVY ( CVY1 ) [4 , 6–8] . The extracellular portion of CrRLK1L proteins shows homology to the animal Malectin protein that has putative carbohydrate binding capacity [24] . The above listed CrRLK1L proteins play roles in diverse environmental contexts , possibly linked to CWI sensing [4 , 6–8] . THE1 , FER and HERK1/2 were found to be required for cell elongation during vegetative growth [25] . FER and ERU have been implicated in polar growth of root hairs [26–28] , and CVY1 was found to control leaf cell morphology and actin cytoskeleton organization [29] . Importantly , FER was recently identified as the receptor for the endogenous peptides RAPID ALKALINIZATION FACTOR 1 ( RALF1 ) and RALF23 that control cell elongation inhibition and immune signaling , respectively [27 , 30] . Furthermore , FER was identified as a key regulator in mechano-sensing , as fer mutant plants show impaired mechanically-induced changes in Ca2+ signalling , transcription and growth [31] . FER was initially implicated in pollen tube reception in the female gametophyte . In fer mutant ovules , pollen tubes do not burst to release the sperm , but instead continue to grow [32–34] . The related ANX1 and 2 are also involved in pollen tube discharge , yet opposite to fer pollen tubes , anx1/2 pollen tubes burst prematurely [35–37] . Finally , fer mutants display enhanced resistance to the powdery mildew Golovinomyces orontii [38] , and the fungus Fusarium oxysporum [39] , which may reflect a role of FER in fungal haustorium formation , while fer mutants are also affected in flg22-induced signalling and are more susceptible to the bacterium Pseudomonas syringae pv . tomato DC3000 [30] . In addition to CrRLK1Ls , another RK subfamily of interest in the context of CWI sensing is the family of wall-associated kinases ( WAKs ) . WAKs can bind pectin [40 , 41] , and WAK1 is involved in the perception of oligogalacturonides ( OGAs ) [42] , which are breakdown products of pectin that can elicit defence responses [43] . In addition , WAKs have been shown to be required for normal cell elongation [44] . Moreover , leucine-rich repeat receptor kinases ( LRR-RKs ) have also been associated with CWI sensing [45] . For example , loss-of-function of the LRR-RK-encoding genes FEI1 and FEI2 results in hypersensitivity to inhibition of cellulose biosynthesis , high sucrose and high salt , and disrupts anisotropic cell expansion and synthesis of cell wall polymers [45] . However , the CrRLK1L THE1 is so far the only RK that was shown to be required for responses to cellulose biosynthesis inhibition . In this study , we expand our understanding of CWI sensing by identifying the recently characterised LRR-RK MALE DISCOVERER 1-INTERACTING RECEPTOR LIKE KINASE 2/LEUCINE-RICH REPEAT KINASE FAMILY PROTEIN INDUCED BY SALT STRESS ( MIK2/LRR-KISS; hereafter referred to as MIK2 ) [46 , 47] as being required for responses to cellulose biosynthesis inhibition . MIK2 shows overlapping as well as distinct functions with THE1 in response to cellulose biosynthesis inhibition . In addition , we find that MIK2 is required for control of normal root growth direction and salt tolerance in a THE1-dependent manner . Moreover , MIK2 plays a role in immunity as it is required for resistance to the fungal root pathogen Fusarium oxysporum . We thus propose that MIK2 is involved in CWI sensing and regulates several aspects of growth , as well as responses to abiotic and biotic stresses .
An overlap exists between responses activated upon disruption of CWI and the ones triggered by perception of microbes [9 , 48] suggesting that CWI signalling and immune signalling might be part of a general ‘danger’ perception system in which loss of CWI would be sensed as ‘altered self’ . Consistently , we observed that treatment with isoxaben ( ISX ) , a chemical widely used to disrupt CWI in a controlled manner via the inhibition of cellulose biosynthesis [1 , 3 , 13] , induced the expression of the genes FRK1 , At1g51890 and CYP81F2 in Arabidopsis , which are commonly used immunity marker genes [49] ( Fig 1A ) . While this increased expression was visible in wild-type Col-0 at 6 and 9 h after treatment , it was absent in the ISX-insensitive mutant ixr1-1 [50] ( Fig 1A ) . Moreover , treatment with other cellulose biosynthesis inhibitors , such as 2 , 6-di-chlorobenzonitrile ( DCB ) [51] and thaxtomin ( TXT ) [52 , 53] , also induced expression of the same genes ( Fig 1B ) . Mild hyper-osmotic stress triggered by mannitol treatment did not activate , but rather seemed to repress the expression of these genes ( Fig 1B ) , revealing that the response observed upon treatment with cellulose biosynthesis inhibitors differs from the response to hyper-osmotic stress . Given the load-bearing role of cellulose in plant cell walls , its loss/reduction may lead to mechanical disruption of cell wall and membrane integrity , the release of cell wall components ( such as carbohydrates or proteins ) , or the active production/secretion of endogenous peptides in response to cell wall damage . Such molecules or mechanical signals might then act as triggers for cell surface RKs . To test this hypothesis , we sought to identify RKs that are required for ISX-induced responses and may therefore represent potential components involved in CWI sensing . Towards this end we tested the ISX response of Arabidopsis T-DNA mutants available in our laboratory with insertions in RK-encoding genes . As a result , we identified two independent homozygous insertion alleles in the gene At4g08850 that displayed reduced ISX-induced immune marker gene expression ( Figs 2A and S1A–S1C ) . This gene encodes a LRR-RK recently characterized as MIK2/LRR-KISS [46 , 47] . We found that mik2-1 was also compromised in DCB- and TXT-induced gene expression ( Fig 2A ) . In addition , mik2-1 was tested for the previously reported ISX-induced JA and SA accumulation , as well as lignin deposition [3 , 9] , and the mutant was found to be impaired in ISX-induced JA accumulation and lignin deposition , but not in ISX-induced SA accumulation ( Fig 2B–2E ) . Together , these data demonstrate that MIK2 is an important regulator of responses triggered by cellulose biosynthesis inhibition . MIK2 contains an extracellular domain consisting of 24 LRRs , a single-pass transmembrane domain , and an intracellular kinase domain ( S1D Fig ) . In accordance with its predicted subcellular localization , MIK2-GFP localized to the plasma membrane ( S1E Fig ) . MIK2 is part of the sub-family XIIb of LRR-RKs [54 , 55] and has a close homolog ( 60% amino acid identity ) , At1g35710 , that we named MIK2-LIKE ( S2A Fig ) . When compared with LRR-RKs encoded by the rice , tomato , poplar , grapevine and soybean genomes , AtMIK2 is more similar to AtMIK2-LIKE than to any of the rice , tomato , poplar , grapevine or soybean sequences [55–58] . On the other hand , in the Brassicaceae species Arabidopsis lyrata and Brassica rapa , MIK2 and MIK2-LIKE paralogs clearly exist ( S2A Fig ) . AtMIK2 and AtMIK2-LIKE are expressed throughout the plant , in young as well as in mature tissues ( S3 Fig ) . To investigate the potential redundant role of MIK2-LIKE in responses to cellulose biosynthesis inhibition , two T-DNA insertion alleles for MIK2-LIKE ( mik2-like-1 and mik2-like-2; S2B and S2C Fig ) , and mik2-1 mik2-like-1 and mik2-1 mik2-like-2 double mutants were tested for ISX-induced responses . Unlike mik2-1 , mik2-like-1 was not impaired in ISX-induced gene expression , JA accumulation or lignin deposition ( S2D–S2F Fig ) . The mik2-1 mik2-like-1 and mik2-1 mik2-like-2 double mutants showed a phenotype similar to the mik2-1 single mutant ( S2D–S2F Fig ) . Thus , despite their close homology , our data suggest that MIK2-LIKE does not fulfil the same function as MIK2 in responses to cellulose biosynthesis inhibition . A prominent CWI sensor candidate is the CrRLK1L THE1 , which is required for cellulose biosynthesis inhibition responses in prc1-1 , a mutant in the cellulose synthase subunit CesA6 [23] . Like MIK2 , THE1 is expressed throughout the plant , in young as well as in mature tissues ( S3 Fig ) . We tested if MIK2 and THE1 play similar roles in responses to cellulose biosynthesis inhibition . We found that both mik2-1 and the1-1 , as well as the double-mutant mik2-1 the-1 were impaired in the ISX-induced expression of the immune marker genes FRK1 and At1g51890 ( Fig 2A ) . However , while mik2-1 and mik2-1 the1-1 were also impaired in the ISX-induced expression of CYP81F2 , the1-1 was not ( Fig 2A ) . Interestingly , immune marker gene expression in response to DCB was also compromised in mik2-1 , the1-1 , and mik2-1 the1-1 ( Fig 2A ) . In contrast , immune marker gene expression in response to TXT was only impaired in mik2-1 and mik2-1 the-1 , but not in the1-1 ( Fig 2A ) , suggesting that MIK2 and THE1 might function in the activation of responses to cellulose biosynthesis inhibition through different mechanisms . More in depth knowledge on the difference between ISX- , and TXT-mode-of-action will however be required to gain further insight in the different mechanisms by which MIK2 and THE1 might operate . ISX-induced JA accumulation was more strongly attenuated in the1-1 and mik2-1 the1-1 than in the mik2-1 single mutant ( Fig 2B ) . ISX-induced SA accumulation was also impaired in the1-1 and mik2-1 the1-1 , but not in mik2-1 ( Fig 2C ) . ISX-induced lignin deposition was impaired to a similar level in mik2-1 , the1-1 , and mik2-1 the1-1 ( Fig 2D and 2E ) . However , unlike THE1 , MIK2 is not required for the cellulose biosynthesis inhibition response in the CesA6 mutant prc1-1 , as loss-of-function of MIK2 did not rescue the shortened dark-grown hypocotyl phenotype in prc1-1 plants , while loss-of-function of THE1 partially did ( S4 Fig ) . In addition to the above described responses , ISX was previously shown to induce rapid internalization of the cellulose synthase complex and accumulation of the complex in microtubule-associated cellulose synthase compartments ( MASCs ) in the cell cortex [59–61] . Neither loss-of-function of MIK2 nor of THE1 interfered with ISX-induced GFP-CESA3 internalization ( S5A–S5C Fig ) , indicating that MIK2 and THE1 must function either downstream , or independent of cellulose synthase complex internalization . In all assays , the mik2-1 the1-1 double mutant displayed the same phenotype as either one of the mik2-1 or the1-1 single mutants ( Fig 2A–2E ) , demonstrating that loss-of-function of both MIK2 and THE1 does not have an additive effect . From a classical genetics point-of-view this would suggest that the two RKs could function in the same pathway; however , clear differences exist in amplitude as well as type of responses that MIK2 and THE1 regulate ( Fig 2A–2E; S4 Fig ) , indicating that they might also regulate different aspects of the CWI maintenance response . It is hypothesized that proper CWI sensing is important for optimal plant growth or development . Interestingly , when grown vertically on MS agar plates , mik2-1 and mik2-2 plants displayed left-ward root skewing , while the1-1 and the1-4 did not ( Fig 3A , S1F Fig , S6A Fig ) . This effect was previously observed in certain Arabidopsis ecotypes , but is minimal in Col-0 [62] . Surprisingly , this effect was abolished in the mik2-1 the1-1 double mutant ( Fig 3A ) . Furthermore , we observed that the presence of ISX or DCB in the growth medium impaired root skewing in mik2-1 ( Fig 3B and 3C ) . The root skewing phenotype of mik2-1 was also attenuated in the prc1-1 genetic background ( Fig 3D ) . Thus , these results indicate that MIK2 controls root angle in a THE1- and cellulose synthase-dependent manner . Although MIK2-LIKE did not fulfil the same function as MIK2 in responses to cellulose biosynthesis inhibition ( S2D–S2F Fig ) , mik2-like-1 and mik2-like-2 displayed a trend towards enhanced root skewing ( S2G Fig ) . However , the enhanced root skewing was only found to be statistically significant in 3 out of 6 experiments . Thus , MIK2-LIKE might contribute to the control of root growth angle , yet not to the same extent as MIK2 . Surprisingly , the mik2-1 mik2-like-1 and mik2-1 mik2-like-2 double mutants displayed a trend towards enhanced root skewing similar to mik2-like-1 and mik2-like-2 single mutants , yet reduced compared to the mik2-1 single mutant . Future work is needed to unravel the genetic relatedness between MIK2 and MIK2-LIKE with respect to control of root growth angle ( S2G Fig ) . To analyse the potential mechanism underlying the root skewing phenotype of mik2 mutants , we investigated if roots of mik2-1 mutants are affected in cellulose microfibril orientation or cell wall structure . Root tips of mik2-1 , the1-1 , and mik2-1 the1-1 , did not display altered cellulose microfibril orientation compared to Col-0 ( S7A Fig ) . Fourier-transform infrared ( FT-IR ) spectroscopy revealed small differences in the cell wall structure in the root tip of mik2-1 plants compared to Col-0 ( S7B and S7C Fig ) . The cell wall structure in the root tips of the1-1 plants was also significantly different from Col-0 , yet showed absorption spectra different from mik2-1 ( S7B and S7C Fig ) , suggesting distinct cell wall modifications . The absorption spectra in the mik2-1 the1-1 double mutant followed a pattern that was more similar to the1-1 than mik2-1 ( S7B and S7C Fig ) , suggesting that the effect of the1-1 on the cell wall is dominant over the effect of mik2-1 . Root tip morphology was comparable between mik2-1 and the1-1 single mutants , and the mik2-1 the1-1 double mutant ( S8 Fig ) . Thus , the distinct influences of mik2-1 and the1-1 on cell wall structure in the root tip might underlie the observed root skewing , or lack thereof , in the mik2-1 single mutant and the mik2-1 the-1 double mutant , respectively . However , biochemical analysis of cell walls from whole roots did not reveal any significant changes in cellulose , hemicellulose or pectin content in the single mutants nor in the mik2-1 the1-1 double mutant ( S9 Fig ) . The observed cell wall defects in mik2-1 and the1-1 are therefore suggestive of subtle , local changes in the root tip , which would need to be confirmed in future , more detailed studies . Recently , natural variation in MIK2 was found to be linked to shoot growth under salt stress conditions in a study in which it was named LRR-KISS [47] . Accessions with MIK2 expression higher than in Col-0 , such as Cen-0 , were less sensitive to salt stress , while accessions with MIK2 expression lower than Col-0 , such as HR-5 , were more sensitive to it [47] . We were thus curious to test the effects of salt stress on mik2 insertional mutant plants in the Col-0 background . In line with a previous report [63] we observed that when grown on MS medium containing 75 mM NaCl , Col-0 roots display a mild skewing response to the right , when seen from the front ( Fig 4A ) . In support with the proposed role for MIK2 in salt stress signalling [47] , mik2-1 plants showed a strongly increased right-ward skewing on medium containing 75 mM NaCl , while not on MS medium containing 150 mM sorbitol ( Fig 4A ) . Unlike mik2-1 , the1-1 and the1-4 were not affected in NaCl-induced changes in root growth direction compared to Col-0 ( Fig 4A , S6B Fig ) . The enhanced NaCl-induced right-ward skewing of mik2-1 roots was abolished in mik2-1 the1-1 roots ( Fig 4A ) . In support with these observations , we found that NaCl-induced reduction in dry weight of mature plants was enhanced in mik2-1 compared to wild-type Col-0 , while the1-1 and the1-4 single and mik2-1 the1-1 double mutants were not affected in the NaCl-induced decrease of dry weight ( Fig 4B , S6C Fig ) . However , of note is that untreated mik2-1 the1-1 plants show a slight reduction in dry weight ( S10 Fig ) , suggesting that loss of both MIK2 and THE1 impairs biomass assimilation under basal conditions . Nevertheless , altogether these data show that MIK2 is required for salt stress tolerance in a THE1-dependent manner . Given that cellulose biosynthesis inhibition leads to the induction of MIK2-dependent responses that are similar to those caused upon perception of microbes or wounding , we were curious to test whether MIK2 could play a role in disease resistance . Interestingly , mik2-1 plants displayed enhanced susceptibility to the root-infecting fungus Fusarium oxysporum isolate Fo5176 ( Fig 5A–5C ) . A similar trend was observed in the1-1 plants , yet was only found to be statistically significant in 4 out of 7 experiments ( Fig 5A–5C , S6D and S6E Fig ) . Mutant the1-4 plants did not display such an enhanced susceptibility phenotype ( Fig 5A–5C , S6D and S6E Fig ) . The mik2-1 the1-1 double mutant plants exhibited a phenotype similar to mik2-1 ( Fig 5A–5C ) . Thus , while MIK2 is required for salt stress tolerance in a THE1-dependent manner , the role of MIK2 in resistance against Fusarium oxysporum isolate Fo5176 does not depend on THE1 . As we obtained discrepant results with the different alleles for THE1 , the exact role of THE1 in resistance to Fusarium oxysporum isolate Fo5176 remains to be elucidated .
In this study , we have identified the LRR-RK MIK2 as an important regulator of responses to cellulose biosynthesis inhibition , as evidenced by the impaired gene expression , JA accumulation and lignin deposition triggered by chemical inhibition of cellulose biosynthesis observed in mik2 mutant plants ( Fig 2 ) . This finding suggests a role for MIK2 in transmission of biochemical or physical signals directly derived from the cell wall or indirectly produced/secreted upon cell wall damage triggered upon cellulose biosynthesis inhibition . In addition , we found that MIK2 plays a role in control of root growth angle ( Fig 3 ) . Different Arabidopsis ecotypes are known to display different degrees of left-ward root skewing , yet the molecular basis of root skewing is not well understood [62 , 64 , 65] . Mechano-sensing , microtubule organization and cell wall composition are suggested to be linked to this phenomenon [62 , 64 , 65] . Mutants in the CrRLK1L FER are impaired in mechano-sensing and display increased right-handed skewing [31] . The hard agar surface of the growth medium imposes a mechanical barrier; the right-ward root skewing in fer might thus be a consequence of impaired mechano-sensing . Moreover , fer mutants are cellulose deficient [66] and this cell wall deficiency could potentially underlie the mechano-sensing defect in fer . Here , loss of MIK2 seems to lead to small , local cell wall defects as well as root skewing ( Figs 3 and S7 ) , suggesting that MIK2 could also be involved in mechano-sensing . Interestingly though , fer mutant roots skew right-ward , while mik2 mutant roots do so left-ward , suggesting that different cell wall defects may translate into different root growth angles . Root skewing has also been previously reported in microtubules mutants [67–69]; however , we could not detect any difference in the orientation of cellulose microfibrils ( S7A Fig ) , which align with the underlying cortical microtubules [59 , 70–73] , indicating that the root skewing phenotype observed in mik2 plants is more complex . Future work should therefore address the molecular mechanisms underlying the observed root skewing . Additionally , we found that mik2 shows increased salt sensitivity ( Fig 4 ) . Mutants with altered cell wall composition or structure were previously shown to display enhanced NaCl sensitivity [74 , 75]; the increased salt sensitivity of mik2 mutants might thus be connected to its cell wall defects . In addition , we observed that mik2 mutants display increased susceptibility to the hemi-biotrophic root pathogen F . oxysporum ( Fig 5 ) , while not to Arabidopsis leaf pathogens , such as the hemi-biotrophic bacterium Pseudomonas syringae pv . tomato DC3000 , the obligate biotrophic oomycete Hyaloperonospora arabidopsidis Noco-2 , or the necrotrophic fungus Plectosphaerella cucumerina BMM ( S11A–S11C Fig ) . In addition , it was previously found that mik2 mutant plants are not affected in resistance against the powdery mildew species Golovinomyces orontii and Erysiphe pisi [76] . We speculate that the role of MIK2 in F . oxysporum resistance is linked to a specific function in the root , which is possibly connected to CWI sensing . Altogether , our results indicate that MIK2 is involved in a diverse array of biological processes in different tissues , similar to the candidate CWI sensor CrRLK1L FER that plays a role in cell elongation , mechano-sensing , pollen tube reception and immunity [8] . In all these processes , feedback information from the cell wall could play a potential important role . It is thus tempting to speculate that these diverse phenotypes of mik2 and fer mutants are linked to a role in cell wall integrity sensing . Up to now , one of the strongest candidate CWI sensors is the CrRLK1L THE1 , as it is so far the only RK that displays impaired responses to cellulose biosynthesis inhibition [23] . FER and other malectin-like CrRLK1L family members have been proposed to play a role in CWI sensing based on the putative carbohydrate-binding capacity of their malectin domains , their structural resemblance to THE1 , and their role in regulation of cell growth in diverse contexts [4 , 6 , 8] . In this study , we compared the phenotype of mik2-1 with that of the1-1 , and found that both RKs are required for responses to cellulose biosynthesis inhibition . However , differences exist in the extent to which these RKs regulate activation of immune marker genes and defence hormone production ( Fig 2 ) , suggesting these RKs might fulfil different functions . However , the function of MIK2 and THE1 seems to be linked , as the left-ward root skewing as well as enhanced salt sensitivity in mik2-1 are abolished in the1-1 genetic background ( Figs 3 and 4 ) . Intriguingly , mik2-1 and the1-1 seem to have distinct effects on cell wall structure in the root tip ( S7 Fig ) , which could potentially underlie the observed root skewing and salt sensitivity in mik2-1 and absence of thereof in mik2-1 the1-1 . Loss of a cell wall sensor disrupts a cell wall-to-cell feedback loop; if such feedback information is lost , one could envision compensatory changes in cell wall composition and properties . Changes in non-cellulosic components can change the physical properties of the cell wall , and might thus affect the interaction between the root surface and the agar ( e . g . the extent to which the root can resist the physical pressure of the agar could be different ) . This could subsequently influence the skewing angle under which the root grows , as well as its responses to external factors . We therefore hypothesize that loss of MIK2 results in mis-regulation of CWI sensing leading to local changes in cell wall composition that impact on root skewing and salt sensitivity . It is tempting to speculate that THE1 is required for these processes through sensing of a ( cell wall-derived ) signal in mik2 . Alternatively , the lack of root skewing and salt sensitivity phenotypes in mik2-1 the1-1 might result from changes in cell wall composition caused by loss of THE1 that overrule changes caused by loss of MIK2 . Of note is that cell wall disruption by inhibition of the cellulose synthase complex interfered with the root skewing response in mik2-1 ( Fig 3 ) , which strengthens the hypothesis that root skewing is connected to cell wall changes . On the other hand , the observed effects of mik2-1 and the1-1 mutations on root growth direction , salt sensitivity , and cell wall structure could be consequences of another , potentially common , underlying cause . To distinguish between the different possibilities , additional insight into the type of cell wall changes that seem to occur in mik2 versus the1 mutant plants could prove useful . However , biochemical analysis of cell walls from whole roots did not reveal any significant changes in cell wall composition in the mutants compared to Col-0 ( S9 Fig ) . The observed cell wall defects in mik2-1 and the1-1 might thus be subtler , local changes in the root tip , and will therefore be more difficult to detect in biochemical analysis . Previously , LRR-RLKs FEI1 and FEI2 have been associated with CWI sensing [45] . However , opposite to mik2 , the fei1 fei2 double mutant shows increased sensitivity to inhibition of cellulose biosynthesis . Moreover , fei1 fei2 is hypersensitive to high sucrose and high salt , and is disrupted in anisotropic cell expansion as well as in the synthesis of cell wall polymers [45] . These findings strengthen the suggestion that responsiveness to cellulose biosynthesis , cell wall composition and salt sensitivity are connected , and form another example of the involvement an LRR-RK in CWI sensing . However , the opposite effects of cellulose biosynthesis inhibition on mik2 mutants compared with fei1 fei2 suggest distinct roles for these proteins in CWI sensing . Interestingly , we found that MIK2 is required for resistance against the root pathogen F . oxysporum , yet this role of MIK2 does not require THE1 ( Fig 5 ) . The effect of THE1 on F . oxysporum resistance seems therefore distinct from its effect on root growth direction and salt sensitivity . The exact role of THE1 in resistance to F . oxysporum remains to be determined , as we found discrepant results with two different alleles ( Fig 5 , S6D and S6E Fig ) . Of note is that the1-4 has recently been suggested to be a gain-of-function , rather than a loss-of-function allele , which might explain the observed discrepancy [77] . Additional alleles would thus need to be tested . If THE1 is involved in resistance against F . oxysporum , MIK2 and THE1 might play a role through separate mechanisms . However , loss-of-function of both MIK2 and THE1 did not have an additive effect ( Fig 5 ) , suggesting that the two RKs could function in the same pathway . The putative role of THE1 in F . oxysporum resistance is clearly distinct from the related CrRLK1L FER , as an Arabidopsis mutant defective in FER has recently been shown to display enhanced resistance to F . oxysporum , most likely because FER is required for the perception of the secreted fungal RALF peptide that contributes to F . oxysporum virulence [39] . Excitingly , MIK2 was recently identified as part of the receptor complex for the female gametophyte-secreted peptide AtLURE1 that functions as a pollen tube attractant [46] . Moreover , mik2 mutant plants displayed defects in male reproductive transmission and pollen tube guidance [46] . AtLUREs are part of a 6 gene-large species-specific cluster of defensin-like genes in Arabidopsis , expressed in the female gametophyte [78] . The Arabidopsis defensin-like gene family comprises 317 members [79] . Although other members of the AtLURE-receptor complex , MIK1 , MALE DISCOVERER ( MDIS ) 1 and MDIS2 , were not found to be involved in responses to ISX or in root skewing ( S12 Fig ) , AtLUREs or related defensin-like peptides might be interesting ligand candidates for MIK2 during CWI , yet their role in CWI remains to be determined . It will be interesting to assess whether such peptides can be secreted/produced in response to cellulose biosynthesis inhibition , activate cellulose biosynthesis inhibition responses , and/or play a role in the control of root growth direction , salt tolerance and F . oxysporum resistance in an MIK2-dependent manner .
All Arabidopsis thaliana lines used in this study were in the Col-0 ecotype genetic background . The following mutants and transgenic lines were used: ixr1-1 [50] , mik2-1 ( SALK_061769 ) , mik2-2 ( SALK_046987 ) , mik2-like-1 ( SALK_112341C ) , mik2-like-2 ( GK-031G02-014862 ) , mik2-1 mik2-like-1 , mik2-1 mik2-like-2 , the1-1 ( outcrossed from prc1-1 the1-1 [23] ) , the1-4 [25] , mik2-1 the1-1 , GFP-CESA3 cesa3je5 [82] , GFP-CESA3 cesa3je5 mik2-1 , GFP-CESA3 cesa3je5 the1-1 , prc1-1 [83] , mik2-1 prc1-1 , the1-1 prc1-1 , mik1 [46] , mdis1-2 [46] , mdis2 [46] , and mdis1-2 mdis2 [46] . The following primers were used for genotyping of mik2-1 , mik2-2 and mik2-like-1: Genotyping of the1-1 , prc1-1 , and cesa3je5 was performed by PCR amplification with the following primers: Next , PCR products were digested with BamHI ( Invitrogen , Carlsbad , CA , USA ) ( cuts THE1 ) , SpeI ( Roche , Basel , Switzerland ) ( cuts the1-1 ) , MfeIHF ( New England Biolabs , Ipswich , MA , USA ) ( cuts PRC1 ) , and HphI ( New England Biolabs ) ( cuts cesa3je5 ) for 4 h at 37°C following manufacturer’s instructions . Digested PCR products were separated on a 3% agarose gel in TBE ( for THE1/the1-1 and PRC1/prc1-1 ) or 1% agarose in TBE ( for CESA3/cesa3je5 ) . The MIK2 coding sequence was amplified from Col-0 cDNA using the primers 5’-CACCATGAACAAAACAAACCCAG-3’ and 5’-AGAAAAGGCAGTGGAGATAGAGAGC-3’ . The corresponding amplicon was cloned into pENTR/D-TOPO using the pENTR Directional TOPO Cloning Kit ( Invitrogen , CA , USA ) . The insert was then transferred into the Gateway-compatible binary vector pEarleyGate103 [84] using GATEWAY LR CLONASE II enzyme ( Invitrogen ) . The final construct was electroporated into Agrobacterium tumefaciens strain GV3101 [85] . For gene expression analysis , seeds were sown on full strength Murashige and Skoog ( MS ) medium ( 4 . 41 g/L; including vitamins; Duchefa , Haarlem , The Netherlands ) and 1% sucrose supplemented with 0 . 8% agar . The seeds were stratified for 2 days at 4°C , and incubated for 5 days at 22°C under a 16-h photoperiod . Seedlings were then transferred to liquid MS medium with 1% sucrose , and grown for another 7 days , after which the growth medium was refreshed . Next day , plants were mock treated , or treated with 0 . 6 μM isoxaben ( ISX ) ( Sigma-Aldrich , St . Louis , MO , USA ) , 6 μM 2 , 6-dichlorobenzonitrile ( DCB ) ( Sigma-Aldrich ) , 0 . 4 μM thaxtomin ( TXT ) ( Sigma-Aldrich ) , or 400 mM mannitol as indicated in the figures . ISX and DCB were added from respectively 1 . 2 mM and 12 mM stocks in DMSO; TXT was added from a 800 μm stock in 100% ethanol . All treatments contained equal amounts of DMSO and ethanol . Total RNA was extracted using Trizol reagent ( Invitrogen ) according to the manufacturer’s instructions . RNA samples were treated with Turbo DNA-free DNase ( Ambion/Thermo fisher Scientific , Waltham , MA , USA ) according to the manufacturer’s instructions . RNA was quantified with a Nanodrop spectrophotometer ( Thermo fisher Scientific ) . cDNA was synthesized from 5 μg RNA using SuperScript III Reverse Transcriptase ( Invitrogen/Thermo fisher Scientific ) according to the manufacturer’s instructions . cDNA was amplified by quantitative PCR using SYBR Green JumpStart Taq ReadyMix ( Sigma-Aldrich ) and the PTC-200 Peltier Thermal Cycler ( Bio-Rad Laboratories , Hercules , CA , USA ) . The relative expression values were determined using U-box as reference and the comparative Ct method ( 2-ΔΔCt ) . The following primers were used for quantitative RT-PCR: Arabidopsis seedlings were grown in liquid culture as described in [9] . Six day-old seedlings were brought into new flasks with growth medium supplemented with either DMSO ( mock ) or 0 . 6 μM ISX . At 7 h after treatment , seedlings were harvested in liquid N2 and JA and SA were extracted and measured as described [87] . At 12 h after treatment , seedlings were harvested in 70% EtOH and stained for lignification using phloroglucinol-HCl as described in [9] . For determination of lignin deposition in the root elongation zone , pictures were taken with a Zeiss Axio Zoom . V16 stereo microscope . Phlorogucinol-stained areas were quantified using ImageJ software and normalized to the total root area photographed , while the root length was kept equal in all images . The ratios obtained are plotted as fold change compared to Col-0 . Seeds were sown on square plates with full strength MS medium ( 4 . 41 g/L; including vitamins; Duchefa ) and 1% sucrose supplemented with 0 . 8% agar . The seeds were stratified for 2 days at 4°C , and incubated for 5 days at 22°C in the dark , in an upright position . Seeds were sown on square plates with full strength MS medium ( 4 . 41 g/L; including vitamins; Duchefa ) and 1% sucrose supplemented with 0 . 8% agar . Where indicated in the figures , growth medium contained DMSO ( mock ) , 2 nM ISX ( Sigma-Aldrich ) , or 25 μM DCB ( Sigma-Aldrich ) . ISX and DCB were added from respectively 80 μM and 1 mM stocks in DMSO . All treatments contained equal amounts of DMSO . The seeds were stratified for 2 days at 4°C , and incubated for 9 days at 22°C under a 16-h photoperiod , in an upright position under a 10° angle relative to the direction of gravity . Seeds were sown on full strength MS medium ( 4 . 41 g/L; including vitamins; Duchefa ) and 1% sucrose supplemented with 0 . 8% agar . The seeds were stratified for 2 days at 4°C , and incubated for 5 days at 22°C under a 16-h photoperiod . Seedlings were then transferred to liquid MS medium with 1% sucrose , and grown for another 2 days , after which the plants were mock treated , or treated with 0 . 6 μM ISX ( Sigma-Aldrich ) for 5 h . ISX was added from a 1 . 2 mM stock in DMSO . Mock and ISX treatment contained an equal amount of DMSO . Seedlings were harvested in 100% ethanol . Root and shoot tissue was separated , 100 roots were used per sample . Root tissue was washed once in ethanol and twice in acetone , and roots were dried overnight . Galacturonic acid content of a Homogalacturonan enriched fraction was determined by incubation of the roots with 100 μL 1% ammonium oxalate ( pH 5 ) for 2 h at 80°C , shaking at 300 rpm . The supernatant was collected , samples were diluted 10 times , and sulfuric acid was added ( 1 . 5 mL sulfuric acid per 250 μL sample in glass tubes ) . Samples were incubated for 15 min at 100°C , kept on ice for 5 minutes . Galacturonic acid content was then measured following the method described in [88] , adapted from [89] . A standard range of galacturonic acid ( 0–0 . 1 g/L ) was included to calculate uronic acid concentration . Cellulose and monosaccharide levels were determined as described [90] . Seedlings were grown and treated as described under “Biochemical analysis of the cell wall” . Seedlings were harvested in ethanol . One day prior to measuring , ethanol was replaced by milliQ water . Seedlings were mounted on gold coated glass slides ( Thermo fisher Scientific ) and dried for 20 min at 37°C . Per root , 20 adjacent areas of 40 μm by 40 μm along the lowest 800 μm of the root , on the side of the central cylinder were selected for spectra collection . Per sample 4 roots were measured , and the experiment was repeated 4 times . Spectra were collected and normalized as described [91] . Statistical analysis was performed using a Student’s T-test with “R” software as described [92] . A . tumefaciens strains carrying MIK2-GFP ( pEarleyGate103/35S::MIK2-GFP-6xHis ) was used for transient expression in N . benthamiana . Transient expression and imaging was realized as described [93] . Cell plasmolysis was induced by treatment with 1 M NaCl for 20 min . Seeds were sown on square plates with full strength MS medium ( 4 . 41 g/L; without vitamins; Duchefa ) and 1% sucrose supplemented with 0 . 8% agar . The seeds were stratified for 2 days at 4°C , and plates were incubated in an upright position for 4 days at 22°C under a 16-h photoperiod . Seedlings were transferred to liquid MS medium with 1% sucrose , and were mock treated , or treated with 0 . 1 μM ISX ( Sigma-Aldrich ) for 2 h . ISX was added from a 0 . 1 mM stock in DMSO . Mock and ISX treatment contained an equal amount of DMSO . GFP-CESA3 was imaged as described previously [94] . Seedlings were grown as described under “Imaging of GFP-CESA3” , yet here seedlings were grown for 7 days . Pontamine Fast Scarlet 4B staining was performed as described in [94] , with some modifications . Seedlings were fixed under vacuum in 4% paraformaldehyde in 0 . 5 X MTSB buffer with 0 . 1% Triton for 1 h . Seedlings were washed in 1 X PBS , and incubated overnight at room temperature in 0 . 003% Pontamine Fast Scarlet 4B ( Sigma-Aldrich ) in 1 X PBS . Next , seedlings were washed with 1 X PBS , mounted in 20 μg/mL citifluor/DAPI , and imaged using the 514-nm laser line of a SP5 confocal laser scanning microscope ( Leica , Solms , Germany ) equipped with an argon laser , as described in [94] . The orientation of cellulose microfibrils relative to the direction of cell elongation was quantified using ImageJ software . Values from 3 independent experiments were combined; per genotype values of 10 roots were collected , and per root a minimum of 12 cells were measured . Imaging of root tip cells stained with propidium iodide was performed as described [95] . The change in root angle in response to salt or sorbitol was determined in seedlings grown on agar plates under a 16-h photoperiod . Plants were germinated on ½ MS medium without sucrose . After 4 days , plants were transferred to new medium with 0 mM or 75 mM of NaCl , or 150 mM of sorbitol ( comparable in osmolarity to 75 mM of NaCl ) . Six days after transfer ( 10-day-old seedlings ) , plates were scanned with an Epson scanner from below . Roots were traced with SmartRoot ( plugin in ImageJ software ) and the directionality output was used to determine the angle of the root ( after transfer ) . The experiment was repeated three times with similar results . For determination of salt tolerance , plants were grown in pots under an 11-h photoperiod , at 22 degrees and 70% humidity . One week after germination , plants were transferred to pots which were saturated with 4 L of either 0 or 75 mM of NaCl solution . During the experiment , all plants were watered with rainwater from below . Conductivity measurements confirmed that salt levels stayed stable during the experiment . After 4 weeks of treatment , plants were cut off and dried in an oven on 68 degrees for 1 week to determine dry weight . Plants were randomised over trays using a randomized block design . Randomisation was similar for each treatment . The experiment was repeated three times with similar results . F . oxysporum ( strain Fo5176; originally isolated by Queensland Plant Pathology Herbarium , Queensland Department of Primary Industries and Fisheries , Brisbane , Australia ) was grown on Czopek-Dox-Agar medium . To obtain spores , an agar plug was added to liquid medium consisting of 3% sucrose , 100 mM KNO3 and 0 , 17% yeast nitrogen base and incubated on a shaker for 3 days . Spores were harvested by filtrating through miracloth , washed and diluted with water . 2-week-old Arabidopsis plants were inoculated by pipetting 750 μL spore solution ( 107 spores/ml ) 1–2 cm deep into the soil , directly next to a plant . Subsequently plants were grown in a climate chamber at 11-hour light/ 13-hour dark cycle , 28°C and 80% relative humidity . The number of chlorotic leaves was counted 12 days post inoculation , and the number of decayed plants estimated 3 weeks post inoculation . Pseudomonas syringae pv . tomato DC3000 infections were carried out on 4-week-old plants . Overnight bacterial culture was pelleted and resuspended in 10 mM MgCl2 to an OD600 of 0 . 02 in presence of 0 . 02% ( v/v ) Silwet L-77 . Bacteria were sprayed onto leaf surfaces , and plants were maintained covered . Two days post-inoculation , leaf discs were sampled and ground in 10 mM MgCl2 . After dilution and plating on Luria-Bertani agar with appropriate selection , plates were incubated at 28°C and colonies were counted 2 days later . P . cucumerina BMM inoculation was carried out on 18-day-old soil-grown plants by spraying a suspension of 4x106 spores/mL of the fungus . Disease progression in the inoculated plants was estimated by an average disease symptom ( 0–5 ) as previously described [96] . Inoculations with spore suspensions of Hyaloperonospora arabidopsidis Noco2 isolate ( 5x104 spores/mL ) were performed on 11-day-old seedlings grown under short day conditions . Progression of the infection was scored after 7 days as previously described [97] . | Plants are constantly exposed to external stresses of biotic and abiotic nature , as well as internal stresses , resulting from growth and mechanical tension . Feedback information about the integrity of the cell wall can enable the plant to perceive such stresses , and respond adequately . Plants are known to perceive signals from their environment through receptor kinases at the plant cell surface . Here , we reveal that the Arabidopsis thaliana receptor kinase MIK2 regulates responses to cell wall perturbation . Moreover , we find that MIK2 controls root growth angle , modulates cell wall structure in the root tip , contributes to salt stress tolerance , and is required for resistance against a root-infecting pathogen . Our data suggest that MIK2 is involved in sensing cell wall perturbations in plants , whereby it allows the plant to cope with a diverse range of environmental stresses . These data provide an important step forward in our understanding of the mechanisms plants deploy to sense internal and external danger . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"cell",
"walls",
"chemical",
"compounds",
"cellulose",
"plant",
"growth",
"and",
"development",
"plant",
"cell",
"biology",
"brassica",
"marker",
"genes",
"organic",
"compounds",
"plant",
"cell",
"walls",
"developmental",
"biology",
"plant",
"science",
"model",
"organisms",
"experimental",
"organism",
"systems",
"seedlings",
"molecular",
"biology",
"techniques",
"plants",
"cellular",
"structures",
"and",
"organelles",
"research",
"and",
"analysis",
"methods",
"arabidopsis",
"thaliana",
"gene",
"expression",
"chemistry",
"molecular",
"biology",
"biochemistry",
"plant",
"cells",
"plant",
"and",
"algal",
"models",
"cell",
"biology",
"organic",
"chemistry",
"root",
"growth",
"genetics",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"physical",
"sciences",
"biosynthesis",
"organisms"
] | 2017 | The Arabidopsis leucine-rich repeat receptor kinase MIK2/LRR-KISS connects cell wall integrity sensing, root growth and response to abiotic and biotic stresses |
9 million people are infected with Trypanosoma cruzi in Latin America , plus more than 300 , 000 in the United States , Canada , Europe , Australia , and Japan . Approximately 30% of infected individuals develop circulatory or digestive pathology . While in underdeveloped countries transmission is mainly through hematophagous arthropods , transplacental infection prevails in developed ones . During infection , T . cruzi calreticulin ( TcCRT ) translocates from the endoplasmic reticulum to the area of flagellum emergence . There , TcCRT acts as virulence factor since it binds maternal classical complement component C1q that recognizes human calreticulin ( HuCRT ) in placenta , with increased parasite infectivity . As measured ex vivo by quantitative PCR in human placenta chorionic villi explants ( HPCVE ) ( the closest available correlate of human congenital T . cruzi infection ) , C1q mediated up to a 3–5-fold increase in parasite load . Because anti-TcCRT and anti-HuCRT F ( ab′ ) 2 antibody fragments are devoid of their Fc-dependent capacity to recruit C1q , they reverted the C1q-mediated increase in parasite load by respectively preventing its interaction with cell-bound CRTs from both parasite and HPCVE origins . The use of competing fluid-phase recombinant HuCRT and F ( ab′ ) 2 antibody fragments anti-TcCRT corroborated this . These results are consistent with a high expression of fetal CRT on placental free chorionic villi . Increased C1q-mediated infection is paralleled by placental tissue damage , as evidenced by histopathology , a damage that is ameliorated by anti-TcCRT F ( ab′ ) 2 antibody fragments or fluid-phase HuCRT . T . cruzi infection of HPCVE is importantly mediated by human and parasite CRTs and C1q . Most likely , C1q bridges CRT on the parasite surface with its receptor orthologue on human placental cells , thus facilitating the first encounter between the parasite and the fetal derived placental tissue . The results presented here have several potential translational medicine aspects , specifically related with the capacity of antibody fragments to inhibit the C1q/CRT interactions and thus T . cruzi infectivity .
Trypanosoma cruzi is the protozoan that causes Chagas' disease [1] , an acute and chronic illness affecting 9 million people in Latin America [2] and causing 50 , 000 deaths per year [3]–[5] . Increasing numbers of infected people have been detected in North America , Europe , Australia , and Japan . Indeed , in the United States , more than 300 , 000 cases have been reported [4] , [6] . It is one of the most important neglected parasitic diseases in the Americas and no safe treatment is available [6] . One third of those infected develops incapacitating circulatory or digestive pathology [4] . Pharmacological treatment of the infection , although effective in some cases , is complicated by the toxicity of the main drugs used ( Nifurtimox and Benznidazole ) [4] , [7] . Therefore , identification and immune intervention on different molecular targets , such as those involved in T . cruzi infectivity and in the parasite capacity to inactivate the complement system , together with conventional pharmacological therapies , may result in synergic or even additive effects . Several T . cruzi surface molecules promote infectivity . Among them gp82 , gp30 , gp35/50 , trans-sialidase , gp85 and calcineurin B , are all metacyclic and tissue culture-derived trypomastigote surface molecules , with Ca+2 signal-inducing activities . They play important variable roles in the parasite attachment to host cells and invasion [8] , [9] . Trypanosoma cruzi calreticulin ( TcCRT ) , a 45 kDa protein [10] , containing the KDEL-Endoplasmic Reticulum ( ER ) retention sequence [11] , [12] , translocates from the ER to the parasite exterior , and strongly inhibits the classical pathway of human complement activation [13] . It also inhibits angiogenesis [14] and tumor growth [15] . Most important , on the parasite surface TcCRT behaves as a potent virulence factor [16] . C1q plays an important role in the in vitro T . cruzi infection of macrophage and fibroblast cell lines , although the parasite and host cell receptors for the complement component were not identified [17] . We have shown that C1 interacts with CRT from parasite and human origins . Thus , TcCRT , differently from the other described parasite surface receptors involved in infectivity , interacts with complement component C1 and utilizes it as an adaptor molecule to recognize host cells [16] , [18] . Thus , translocation of TcCRT from the ER to the membrane , not only inhibits the classical pathway of complement by interacting with C1 ( q , r2 , s2 ) [13] , [19] but , in a parasite apoptotic mimicry effort , it also promotes infectivity , most likely by generating effective C1q-mediated “eat me” signals . Attempts to interfere with the C1/TcCRT interactions with whole Igs or their F ( ab′ ) 2 fragments have opposite and predictable outcomes , both in vitro and in vivo . In vitro , in a cell-free system , TcCRT binds C1q , and whole IgG anti-TcCRT mediates Fc-dependent incorporation of additional C1q molecules onto the immune complexes , with likely consequent increased in vivo infectivity [16] . On the other hand , F ( ab′ ) 2 fragments from anti-TcCRT IgGs , devoid of their C1q-fixing Fc domains , revert the TcCRT/C1 interaction [16] . Thus , in mice , whole IgG anti-TcCRT and their F ( ab′ ) 2 fragments respectively stimulate and inhibit T . cruzi infectivity [14] , [16] . Within the uterus , during mammalian gestation , a balance between tolerance to a hemiallogeneic fetus and protection against mother-borne pathogens must be operative . Subversion of this equilibrium by pathogens can complicate pregnancy or lead to vertical transmission of pathogens with fetal , perinatal or later morbidity or mortality [20] . The placenta is a chimeric organ made of maternal and fetal cells that nourishes and protects the fetus . Fetal derived invasive extravillous trophoblasts anchor the placenta in the uterine implantation site ( decidua ) and restructure maternal arteries to facilitate blood access to fetal derived syncytium villous trees [21] . The general consensus is that the syncytiotrophoblast ( ST ) is a formidable barrier to infection with microbes important during pregnancy . These properties of the ST derive from: i ) . Absence of intercellular junctions; ii ) . Absence of E-cadherin [22] , [23]; iii ) . Presence of a network of profuse branched microvilli; iv ) . Presence of a dense cytoskeleton network [24]; v ) . A prevalence of an apical to basal directionality of nutrient transport [25] that may also preclude endocytic uptake of pathogens on the basal side; vi ) . Abundance of fused mitochondria [26] and , vii ) . ST production of reactive nitrogen species [27] . Despite the effectiveness of the placental barrier , mother-to-child transmission leading to congenital Chagas' disease and other adverse neonatal outcomes is increasingly recognized [28]–[31] . During pregnancy , the rate of vertical transmission ofT . cruzi infection is approximately 5–10% ( close to that reported for untreated HIV/AIDS [32] ) . There are over 14 , 000 cases of congenital Chagas' disease now reported in Latin America [33] , with 2 , 000 newborns infected annually in North America alone , a situation also increasing in other developed countries [31] . We have previously shown that T . cruzi induces ST destruction and detachment in human placenta chorionic villi explants ( HPCVE ) , together with selective disorganization of the basal lamina and of collagen I in the connective tissue of the villous stroma [34] , as well as apoptosis [35] , [36] . These effects may be mediated by cruzipain or metalloproteases that degrade the local extracellular matrix components such as collagen type I , IV and fibronectin [37] . Thus , the parasite overcomes the placental barrier and accesses the fetus [34] , [37] . There are important differences between human and mouse placenta [38] , [39] that limit the utility of these animals as experimental models for basic in vivo studies on congenital transmission of this infection; the main difficulties being the low yield of congenital transmission to the litter using different strains [40] , [41] and the fact that murine placenta has a labyrinthine structure and the human placenta a villous one [42] . The use of HPCVE allowed us to propose that , because CRTs from parasite and fetal cell origins interact with maternal complement component C1 , these three molecules strongly promote T . cruzi infection of human placenta . The results presented here agree with this proposal .
Written informed consent for the experimental use of the placenta was given by each patient . The protocols involving the use of human placenta were approved by the Ethics Committee for Research in Human Beings of the Faculty of Medicine , University of Chile ( N° 041-2011 ) and by the Committee for Bioethics from The National Council of Scientific and Technologic Research ( CONICYT-Chile ) . The use of rabbits for the generation of antibodies has been described previously [43] and followed the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health , U . S . A . The protocols were approved by the Bioethics Committee , Faculty of Medicine , and University of Chile and by The National Council of Scientific and Technologic Research ( CONICYT-Chile ) . Animals were maintained in our Central Experimental Animal Facility and were cared for by trained personnel and veterinarians . Human term placentas were obtained from uncomplicated pregnancies from vaginal or caesarean delivery from the Maternity Section , Hospital San José , Santiago , Chile . Exclusion criteria for patients were any maternal , fetal or placental pathology . Placentas were collected in cold PBS and processed no more than 30 min after delivery . Their maternal and fetal surfaces were discarded and villous tissue was obtained from the central part of the cotyledons . HPCVE were washed with PBS in order to remove blood , cut in approximately 0 . 5 cm3 pieces and co-cultured with infective Y strain trypomastigotes ( 2×104/ml ) for 2 h in 1 ml of RPMI . In order to determine the roles of TcCRT , fetal HuCRT and maternal C1q ( the complement moiety deprived of serine proteases C1r and C1s activities ) , in T . cruzi infectivity in HPCVE , the explants were co-cultured with combinations of the following reagents: i ) . C1q ( Complement Technologies , Taylor , Texas , USA ) , ii ) . IgG polyclonal antibodies anti-recombinant TcCRT [43] or anti-recombinant HuCRT [44] , both antisera generated in rabbits in our laboratories , by conventional methodology . Each antisera was tested in western blottings against the homologous and orthologous recombinant antigens , and also against a wild type E . coli extract . As usually occurs with polyclonal antisera , only at high concentrations , both of them marginally recognized the orthologous antigen . No antigen recognition was evident in E . coli extracts [43] , iii ) . F ( ab′ ) 2 IgG fragments anti-recombinant TcCRT or anti-recombinant HuCRT , generated by pepsin digestion [16] , [43] and , iv ) . Fluid phase recombinant HuCRT . ( DNA coding for HuCRT was kindly donated by Prof . Wilhelm Schwaeble , Leicester University , UK , and was expressed and purified in our laboratory ) . HPCVEs were incubated with trypomastigotes in RPMI supplemented with HIFBS ( control ) and C1q , in the presence of HuCRT , in solution . All these reagents were tested in several concentrations , in the different experimental groups , as indicated in the Results section . Basal infectivity was obtained when HPCVE were incubated with trypomastigotes supplemented with heat-inactivated fetal bovine serum ( HIFBS ) . The amount of parasite DNA in HPCVE was determined by qPCR . Experiments shown are representative of those performed in three different placentas . Genomic DNA was extracted from the placental tissue with the Wizard Genomic DNA Purification Kit ( Promega , USA ) , according manufacturer's instructions and quantified by μDrop Plate DNA quantification system in a Varioskan Flash Multimode Reader ( Thermo Scientific , USA ) . For amplification of human and parasite DNA , two specific primer pairs were used . A 100 bp human GAPDH sequence was amplified using the primers hGDH-F ( 50-TGATGCGTGTACAAGCGTTTT-30 ) and hGDH-R ( 50-ACATGGTATTCACCACCCCACTAT-30 ) , designed using the Primer Express software ( version 3 . 0; Applied Biosystems ) . For T . cruzi DNA detection a 182 bp sequence of satellite DNA was amplified by using TCZ-F 50-GCTCTTGCCCACAMGGGTGC-30 and TCZ-R 50-CAAGCAGCGGATAGTTCAGG-30 primers [45] . Relative quantification analysis of the results was expressed as RQ value by the comparative Control ( ΔΔCt ) method [46] . The placental villi were fixed in 10% formaldehyde in 0 . 1 M phosphate buffer ( pH 7 . 3 ) for 24 h , then dehydrated in alcohol , clarified in xylene , embedded in paraffin , and sectioned at 5 µm ( Microtome Leitz 1512 ) . Paraffin histological sections were stained with hematoxylin-eosin for routine histological analysis . In order to detect fetal and maternal CRT in human placenta tissues , standard immunohistochemical techniques were used . Briefly , histological sections were treated with 3% hydrogen peroxide in methanol for 10 minutes and incubated for 30 minutes in Dako Target Retrieval ( Dako , Carpinteria , CA , USA ) on a steamer . The tissue was probed with rabbit anti-human CRT IgG [43] , followed by goat anti-rabbit IgG conjugated to peroxidase and then a commercial substrate ( Histomouse MAX-AEC Broad SpectrumTM Kit ( Invitrogen , Camarillo , CA , USA ) employed , for staining of CRT . Results are expressed as mean and SDs . The significance of differences was evaluated using ANOVA followed by Dunnett's post-test . The control group corresponds to HPCVE incubated with culture media supplemented with heat-inactivated FBS and infected with the parasite in the absence of any of the molecules tested .
To address whether C1q ( the complement moiety deprived of serine proteases C1r and C1s activities ) is involved in the infection of HPCVE , explants were incubated with trypomastigotes , in the alternative presence of C1q or C1q plus whole IgG anti-TcCRT . Increases in parasite DNA of 4 . 5 , and 6 . 0 fold , were respectively observed , as compared to the basal infection in the presence of HIFBS ( Figure 1A ) . Fc-dependent binding of additional C1q molecules could explain why the presence of polyclonal whole IgGs anti-TcCRT increased parasite infectivity [16] . When FBS was used , infectivity was moderately increased most likely because of the presence of residual active C1q , or other non-identified factor ( s ) . For comparison purposes infectivity in the presence of HIFBS was considered as basal and used as control in all subsequent experiments . Since , in order to promote infectivity , C1q binds to trypomastigote translocated CRT , this interaction should be prevented by bivalent TcCRT-binding antibody fragments deprived of their Fc domains and hence unable to bind C1q . Thus , in HPCVE incubated with trypomastigotes , in the presence of C1q and F ( ab′ ) 2 polyclonal antibody fragments anti-TcCRT , the C1q-dependent parasite ability to infect placental cells was reverted to basal levels ( Figure 1B ) . Human placental tissues are known to express high levels of HuCRT [47] . In order to define if ST expresses this molecule , as a possible receptor for TcCRT-bound C1q on the parasite surface , we compared HuCRT expression at the basal maternal decidua and free fetal villi . Polyclonal antibodies against HuCRT readily detected the human chaperone molecule mainly on fetal placenta villi ST , with a distribution consistent with its exposure on the ST surface ( Figure 2 ) . We then aimed at defining whether HuCRT participates as a possible receptor for the C1q/TcCRT complex present on the parasite surface . HPCVE were incubated with trypomastigotes , in the presence of C1q . In Figure 3A fluid-phase HuCRT inhibits even the basal infection ( down to 12% of the control ) , most likely by competing with residual bovine C1q present in the HIFBS . In this experiment the C1q-mediated infectivity again reaches 4 times over the control ( Figure 1A ) . Figure 4B summarizes the results of a fluid-phase HuCRT dose-response capacity to inhibit C1q-mediated HPCVE T . cruzi infection . Complete dose-dependent blocking of the C1q-mediated parasite ability to infect placental cells , is observed . In other words , in the absence of fluid-phase HuCRT , C1q mediates 6–30-fold increase in the T . cruzi capacity to infect HPCVE ( Figure 3 A–B ) . As an infectivity promoter , C1q should simultaneously bind to both HuCRT , on placental ST , and TcCRT , on the parasite . Concurringly , F ( ab′ ) 2 antibody fragments , derived from polyclonal IgG generated against recombinant CRTs from both human and parasite origins , completely reverted the C1q-mediated T . cruzi infectivity in HPCVE . The presence of F ( ab′ ) 2 IgG fragments , obtained from preimmune sera did not alter the infectivity mediated by the complement component ( Figure 4 ) . T . cruzi infection induces ST detachment in chorionic villi , together with disorganization of the basal lamina and of collagen I in the connective tissue [34] . Since the TcCRT/C1q/HuCRT interaction is involved in infectivity , we asked whether these interactions could be intervened , with consequent improvement of the histopathological alterations . Figure 5 summarizes the results obtained when HPCVE were incubated with trypomastigotes in RPMI , supplemented with HIFBS , where a slight detachment of trophoblast from the basal lamina is observed ( Figure 5B , arrowheads ) , as compared with the non-infected control ( Figure 5A ) . In the presence of exogenous human C1q , a strong trophoblast detachment ( arrowheads ) and destruction of fetal connective tissue can be observed ( Figure 5C , arrows ) . Blocking of HuCRT by F ( ab′ ) 2 IgG fragments anti-HuCRT ( Figure 5D ) , or blocking parasite-attached C1q with soluble HuCRT , results in partial prevention of trophoblast detachment ( Figure 5E ) .
The placental barrier is effective to protect the fetus from mother borne microorganisms [48] , but T . cruzi , like several other pathogens , has developed means to trespass it [49]–[57] and the relative relevance of congenital transmission of this infection is increasing [58] , [59] . Several T . cruzi surface molecules promote infectivity [8] , [9] , however their association with T . cruzi placental infection has not been established . Efforts to define molecular mechanisms explaining how T . cruzi trespass the placental barrier are thus important . We have proposed that among the many functions that parasite exteriorized TcCRT displays in the host/parasite interplay [18] , of central relevance is its capacity to interact with classical complement component C1q . Based solely on this property , TcCRT behaves as a main virulence factor [16] . On the host cell side , fetal cC1qR ( membrane-bound CRT acting as a receptor for the collagenous C1q tails ) [60] interacts with TcCRT-bound maternal C1q and thus with the parasite . Based on the known capacity of C1q to interact with Fc domains of IgG [16] , attempts to interfere with the C1/TcCRT interactions with whole Igs or their Fc-deprived F ( ab′ ) 2 fragments have drastically different predictable outcomes , both in vitro and in vivo . In vitro , in a cell-free system , TcCRT binds C1q and F ( ab′ ) 2 fragments from anti-TcCRT IgGs , devoid of their C1q-fixing Fc domains , revert this interaction [43] . Most important , although pretreatment of trypomastigotes with C1q increased infectivity in the RAW murine cell line , as well as mice mortality and parasitemia , the F ( ab′ ) 2 fragments anti-TcCRT significantly interfered with the C1q-dependent infectivity [16] . As a consequence of infection , several phenomena that may facilitate , at least partly , the parasite progress towards fetal tissues have been described: i ) . Disassembly of cortical actin cytoskeleton [59] , [61] , ii ) . Presence of parasite DNA [62] and antigens [34] on the fetal side of placenta , iii ) . ST destruction and detachment in the chorionic villi , disorganization of the basal lamina and collagen I in the connective tissue of the villous stroma ( VS ) [35] , [36] , iv ) . Apoptosis in chorionic villi , especially at the ST and cytotrophoblast [35] , v ) . Parasite cruzipain degradation of extracellular matrix ( ECM ) collagen type I , IV and fibronectin [63] and , vi ) . Endogenous ECM metalloproteases - mediated tissue damage [36] , [64] , [65] . These alterations occur after parasites contact with the HPCVE . Most likely , on the parasite side , translocated TcCRT and , on the maternal side , C1q and fetal HuCRT ( cC1qR ) , all play important roles in the early first contact between trypomastigotes and placental tissue . This will be followed by an elaborate infective process , as described above . There are precedents of in vitro or ex vivo systems to study T . cruzi infectivity of placental tissues [34]–[36] , [62] , [66] . Two constraints to these models have been posed in the literature [67]: i ) . The parasite DNA on the fetal side of placenta may correspond to debris transferred from the maternal side and , ii ) . The frequent use of large numbers of parasites ( i . e . 105–107 for a few mg of chorionic villi ) , is highly superior to what can be expected in chronically infected women with parasitemias below 15 pg/ml . HPCVE assays have been recognized as valid correlates not only to study the tissue damage caused by T . cruzi during placenta infection , but also to explain the earlier stages of vertical transmission [62] , [68] , [69] . First , damage to the tissue was evident when 2×104 parasites were used per each HPCVE assay ( Figure 5 ) and the parasite presence in chorionic villi is also microscopically clear in a similar experimental set up [34] . ( This type of damage has also been reported for cytomegalovirus [70] , Plasmodium falciparum [71] and Toxoplasma gondii [72] . Second , with regard to the large number of parasites used in other reported assays , we calibrated conditions for HPCVE infection down to 2×104 trypomastigotes per each explant with adequate signals for parasite DNA detection in qPCR . This parasite concentration is within the range of expected parasite numbers reaching the placenta of infected women , in a 24 hour period [34] . Although it could be proposed that detachment and destruction of the infected tissue that interacts with the maternal blood , is a mechanism to avoid congenital infection [21] , this mechanism is not efficient enough to mediate sterile fetal protection . In this report we aimed at defining whether the above T . cruzi - infection promoting mechanisms are operative at the placental level . In HPCVE ex vivo assays , at least a 3–4 fold increase in parasite infectivity ( average of the five repetitions shown in Figure 1–3 ) was mediated when both TcCRT and C1q were present . The infection-promoting capacity of FBS and , to a lesser extent , of HIFBS , could be respectively explained by the presence of putative bovine C1q , its active remnants , or other non-identified factor ( s ) . For these reasons , and for comparison purposes , infectivity in the presence of HIFBS was considered as basal and used as control in all subsequent experiments . In placenta , Fc-dependent binding of additional exogenous C1q molecules could explain why the presence of polyclonal IgGs anti-TcCRT also mediated increased parasite infectivity ( Figure 1A ) [16] . We then reasoned that since , in order to promote infectivity , C1q must bind to parasite translocated TcCRT , this interaction should be interfered by F ( ab′ ) 2 polyclonal antibody fragments anti-TcCRT . Accordingly , complete reversion of C1q-dependent infectivity to basal levels was observed ( Figure 1B ) . Teleologically , it could be proposed that , during the co-evolution of the host/parasite interplay , the parasite did develop ways to subvert the humoral immune response against TcCRT . Indeed , if not all , most of humans infected with T . cruzi have IgGs anti-TcCRT in their plasma [73] and , in an apparent paradox , immunization of mice with TcCRT increases parasite infectivity [16] . Moreover , treatment of parasites with whole IgGs anti-TcCRT increases their capacity to infect murine macrophages in vitro . Opposite results are obtained in vivo when the corresponding F ( ab′ ) 2 IgG fragments were passively administered to infected animals , or in vitro when the parasites were treated with these modified antibodies [16] . High levels of HuCRT are known to be expressed in human placental tissues [47] . In order to define if this molecule is expressed at the ST , as a possible receptor for C1q bound to TcCRT on the parasite surface , we compared CRT expression at the maternal basal decidua and in fetal free villi ( Figure 2B ) . HuCRT from maternal and fetal origins were respectively detected at both the decidua and villi . In villi , intense reactivity was detected at the ST , with a distribution consistent with a possible exposure on the ST surface . At the villi stroma the reactivity was weak ( Figure 2B ) . It is thus likely that this fetal CRT may serve as a receptor for C1q already bound to the parasite by means of TcCRT . In other words , we propose that TcCRT-C1q complexes remain on the parasite surface and that C1q bridges the parasite with the ST , as a preamble to infection . This possibility was tested by incubating HPCVE with trypomastigotes , in the simultaneous presence of C1q and increasing concentrations of fluid-phase HuCRT . Complete dose-dependent blocking of the C1q-mediated parasite ability to infect placental cells , is observed . In other words , it is highly likely that fetal HuCRT , at the ST level , binds maternal C1q already bound to TcCRT on the parasite surface ( Figure 3 ) . Moreover , the previous findings were further corroborated by showing that F ( ab′ ) 2 antibody fragments , derived from polyclonal IgG generated against recombinant CRTs from both human and parasite origins , completely and specifically reverted the C1q-mediated T . cruzi infectivity in HPCVE ( Figure 4 ) . Therefore , these antibody fragments effectively blocked the capacities of both ST fetal and parasite CRTs from binding C1q . Thus the infectivity mediated by this complement component was neutralized by these modified antibodies . About 50% overall homology exists between TcCRT and HuCRT ( up to 70% in some domains ) [12] . However , at high concentrations , marginal cross reactivity of the anti- TcCRT antibodies with HuCRT and vice versa is observed in ELISA and IWB assays ( not shown ) . However , this does not affect our conclusion with regard to CRT ( from host and parasite origins ) involvement in infectivity in HPCVE . The use of fluid phase HuCRT to inhibit infectivity , in a dose-dependent manner ( Figure 3B ) , additionally supports this proposal . Consistent with the notion that the parasite CRT/maternal C1q/fetal CRT interactions are involved in infectivity , when HPCVE were incubated with trypomastigotes in RPMI , supplemented with HIFBS , slight detachments of trophoblast from the basal lamina are observed , as compared with the non-infected control . When exogenous human C1q was present , trophoblast detachment was more evident . Blocking fetal CRT by F ( ab′ ) 2 IgG fragments anti-HuCRT , or blocking parasite-attached C1q with soluble HuCRT , resulted in partial prevention of trophoblast detachment ( Figure 5 ) . This is most likely due to decreased parasite penetration into the villi tissue and associated reported damages [34] . Since both Ficolins and MBL also interact with CRT , from human and parasite origins [13] , [74] , they will also probably facilitate parasite infectivity in this experimental set up . Since the “danger” signals detected by these components are different from IgGs , both whole IgGs anti-TcCRT as well as their F ( ab′ ) 2 fragments should inhibit infectivity mediated by these lectin pathway complement components . In vivo models to validate our ex vivo results are complex to implement . Although the murine model , has several advantages to study congenital diseases ( short pregnancy , large litters , short weaning time ) , the rate of T . cruzi vertical transmission to the fetuses is extremely low [75]–[77] and the structure of murine placenta is very different [38] , [39] . TcCRT is a virulence factor as originally proposed by us [16] , [18] . Recently , TcCRT has also been involved in the binding of thrombospondin-1 ( TSP-1 ) , with enhanced infectivity of mouse fibroblasts [78] . Differently from C1q ( and possibly from Ficolins and MBL ) , TSP-1 is an ubiquitously located molecule , capable of interacting with a wide array of cellular proteins [79] . TSP-1 is expressed in fetal villous tissue [80] and could explain , at least partly , the basal T . cruzi infectivity obtained when HPCVE were incubated with trypomastigotes , in the absence of exogenous C1q ( Figures 1; 3–4 ) . Considering the previous results altogether , in Figure 6 we propose a simplified integrated model on the participation of maternal C1q and fetal CRT , on the one side , and parasite CRTs , on the other , in T . cruzi infection of human placenta . Most likely , in vivo , infective T . cruzi trypomastigotes circulate with maternal C1q already bound to translocated TcCRT . The results presented here have several potential translational medicine aspects , specifically related with the capacity of antibody fragments to inhibit the C1q/CRT interactions and thus T . cruzi infectivity . Finally , based on these observations , it could be proposed that in pre immunized mothers , carrying whole anti-TcCRT antibodies , the parasite would be in a better position to infect the fetus . This could be a frequent event , given the high prevalence of anti-TcCRT antibodies in infected humans [73] . | The Trypanosoma cruzi protozoan infects 9 million people in Latin America and increasing numbers in North America , Europe , Australia , and Japan . It is an important neglected parasitic disease in the Americas with no safe treatment available . One third of those infected develops incapacitating pathology . While in poor countries transmission of the parasite is mainly through blood feeding insects , transplacental infection is increasingly important in developed regions . Herein we show that T . cruzi calreticulin ( TcCRT ) , a multifunctional protein , exteriorized by the parasite , mediates infection of human placenta , since it binds human complement component C1 , a “danger signal” detector . ( Complement is an innate immune defense system , with more than 40 plasma or membrane-bound proteins ) . However , in a parasite strategy , maternal C1 is utilized to infect placenta . Fetal calreticulin ( HuCRT ) is also easily detectable in placental tissues that are in direct contact with maternal blood . Thus , C1q by bridging parasite and HuCRT mediates high increases in cultured placental tissue infection with damaging consequences . Complete reversion of C1-mediated infection and a decreased placental damage , is observed in the presence of anti-TcCRT and anti-HuCRT antibody fragments , or fluid-phase competing HuCRT . It remains to be determined whether these mechanisms also operate in other intracellular protozoa . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"medicine",
"infectious",
"diseases",
"chagas",
"disease",
"neglected",
"tropical",
"diseases",
"immunology",
"biology",
"immunomodulation"
] | 2013 | The Interaction of Classical Complement Component C1 with Parasite and Host Calreticulin Mediates Trypanosoma cruzi Infection of Human Placenta |
Our current understanding of how the brain segregates auditory scenes into meaningful objects is in line with a Gestaltism framework . These Gestalt principles suggest a theory of how different attributes of the soundscape are extracted then bound together into separate groups that reflect different objects or streams present in the scene . These cues are thought to reflect the underlying statistical structure of natural sounds in a similar way that statistics of natural images are closely linked to the principles that guide figure-ground segregation and object segmentation in vision . In the present study , we leverage inference in stochastic neural networks to learn emergent grouping cues directly from natural soundscapes including speech , music and sounds in nature . The model learns a hierarchy of local and global spectro-temporal attributes reminiscent of simultaneous and sequential Gestalt cues that underlie the organization of auditory scenes . These mappings operate at multiple time scales to analyze an incoming complex scene and are then fused using a Hebbian network that binds together coherent features into perceptually-segregated auditory objects . The proposed architecture successfully emulates a wide range of well established auditory scene segregation phenomena and quantifies the complimentary role of segregation and binding cues in driving auditory scene segregation .
We live in busy environments , and our surrounds continuously flood our sensory system with complex information that needs to be analyzed in order to make sense of the world around us . This process , labeled scene analysis , is common across all sensory modalities including vision , audition and olfaction [1] . It refers to the ability of humans , animals and machines alike to parse the mixture of cues impinging on our senses , organize them into meaningful groups and map them onto relevant foreground and background objects . Our brain relies on innate dispositions that aid this process and help guide the organization of patterns into perceived objects [2] . These dispositions , referred to as Gestalt principles , inform our current understanding of the perceptual organization of scenes [3 , 4] . In most theoretical accounts , the role of Gestalt principles in parsing a scene can be conceptualized in two stages: segregation ( or analysis ) and grouping ( or fusion ) [5] . In the first stage , the sensory mixture is decomposed into feature elements , believed to be the building blocks of the scene . These features reflect the physical nature of sources in the scene , the state and structure of the environment itself , as well as perceptual mappings of these attributes as viewed by the sensory system . These features vary in complexity along a continuum from basic attributes ( e . g . edges or frequency components ) to more complex characteristics of the scene ( e . g . shapes or timbral profiles ) . The ubiquitous nature of these profiles often conceals the multiplexed structures that underlie this analysis of scene features in the brain . In most computational accounts , this segregation stage is modeled using feature analyses which map the sensory signal into its building blocks ranging from simple components ( e . g . frequency channels ) to dimensionally-complex kernels [6 , 7] . Processing the distinctive features of a scene is generally followed by a fusion stage which integrates the state and behavior of the scene’s building blocks using grouping mechanisms that reflect the local and global distribution and dynamics of the features . This stage employs ‘rules’ that guide how grouped elements give rise to perceptually coherent structures forming objects or streams [2 , 8 , 9] . In many mathematical models , these grouping cues are often leveraged in back-end classifiers that are tuned to capture patterns and relationships within specific object classes ( e . g . speech , music , faces , etc ) [10–13] . In doing so , these models effectively capture the inter-dependencies between object attributes and learn their mapping onto an integrated representational space [14–16] . Ultimately , success in tackling scene analysis depends on two key components [17]: ( i ) obtaining a rich and robust feature representation that can capture object specific details present in the scene; ( ii ) grouping the feature elements such that their spatial and temporal associations match the dynamics of objects within the scene . Vision models have been very successful in mining these two aspects of scene analysis . Intricate hierarchical systems have leveraged inherent structure in static and dynamic images to extract increasingly elaborate features from a scene that are then used to segment it , interpret its objects or track them over time [18–20] . Data-driven approaches have shown that high dimensional feature spaces are very effective in extracting meaningful semantics from arbitrary natural images [20–22]; while hand-engineered features like scale-invariant feature transform ( SIFT ) [23] , histogram of oriented gradients ( HOG ) [24] , and Bag-of-visual-word descriptor [25] among others have also enjoyed a great deal of success in tackling computer vision problems like image classification and object detection . Recent advances in deep layered architectures have resulted in a flurry of rich representational spaces showing selectivity to contours , corners , angles and surface boundaries in images [26–29] . The deep nature of these architectures has also led to a natural evolution from low-level features to more complex , higher-level embeddings that capture scene semantics or syntax [30 , 31] . In audition , computational approaches to tackle auditory scene organization have mostly taken advantage of physiological and perceptual underpinnings of sound processing [17] . A large body of work has built on knowledge of the auditory pathway , particularly the peripheral system to build sophisticated analysis models of auditory scenes . These systems extract relevant cues from a scene , such as its spectral content , spatial structure as well as temporal dynamics; hence allowing sound events with uncorrelated acoustic behavior to occupy different subspaces in the analysis stage . These models are quite effective in replicating perceptual results of stream segregation especially using simple tone and noise stimuli [32–37] . Some models also extend beyond early acoustic features to examine feature binding mechanisms that can be used as an effective strategy in segregating wide range of stimuli from simple tone sequences to spectro-temporally complex sounds like speech and music [38–40] . In most approaches however , the models are built around hand-crafted feature representations , hence limiting their scope to specific mappings of the acoustic space . With the emergence of deep belief architectures , recent efforts started learning rich feature spaces from natural soundscapes in a data driven fashion , and subsequently using these spaces in domains like music genre classification , phoneme classification and speaker identification [41–44] . Applications of deep learning have also successfully tackled the problem of speech separation even with monaural inputs by learning embeddings of a speaker’s time-frequency dynamics against other speakers [45 , 46] . The current study also leverages neural network theory to ‘learn’ Gestalt principles directly from sound . The work examines what kind of cues can one infer from natural sounds; how well do these learned cue reflect the known Gestalt components of auditory streams; and how effective are these cues in explaining perceptual organization of auditory scenes with varying degrees of complexity . The model is devised as a hierarchical structure that generally follows the two-stage pipeline of analysis then fusion , in line with prototypical scene analysis theories [5] . This system analyzes the incoming acoustic signal with a multitude of granularities , hence allowing both local and global acoustic attributes to emerge . The short-term analysis performs a local tiling of the spectro-temporal space; hence inferring simultaneous grouping cues [47–49] . A longer-range analysis extends the segregation stage to examine temporal dependencies across acoustic attributes over different time scales; hence exploring emergence of sequential grouping cues [50–54] . Finally , a fusion stage binds the cues together based on how strongly they correlate with each other across multiple time scales . This integration is achieved using Hebbian learning which reinforces activity across coherent channels and suppresses activity across incoherent ones [55–57] . Apart from the basic layout and choice of analysis window sizes , the network is trained in an unsupervised fashion on a rich sound dataset including speech and nature sounds hence offering a general inference architecture of auditory Gestalt cues that are common across many sound environments . The overall system is tested with a wide range of stimuli where we can quantify the role of each and every component of the network in driving stream segregation processes . We also contrast the system performance with a set of control experiments where different components of the model are deliberately switched on/off in order to examine their impact on the organization of different acoustic scenes . These control experiments aim not only to dissect the role of various system components . They also shed light on how necessary and/or sufficient different grouping cues are to anchor the analysis of different stimuli structures and sound types . The paper first presents an in-depth description of the proposed architecture , followed by an analysis of the emergent properties of the trained network and their potential neural correlates in the auditory pathway . The experimental results outline how the network replicates human psychoacoustic behavior in stream segregation and speech intelligibility paradigms . Finally , we present control experiments that dissect the network architecture and examine the contribution its component . We discuss the implications of this network in shedding light on ties between observed perceptual performance in various complex auditory scenes and the neural underpinnings of this behavior as implemented in networks of neurons along the auditory pathway .
A number of Gestalt principles have been posited as indispensable anchors used by the brain to guide the segregation of auditory scenes into perceptually meaningful objects [8 , 47 , 58] . These comprise a wide variety of cues; for instance harmonicity which couples harmonically-related frequency channels together , common fate which favors sound elements that co-vary in amplitude , and common onsets which groups components that share a similar starting time and to a lesser degree a common ending time . Most of these cues are thought to be innate in our auditory system , and evidence for their role is found across many species [59–63] . These processes likely take advantage of statistical regularities of sounds in natural environments and reflect the physical constraints of sound generation and propagation ( e . g . two sound sources rarely start at the exactly the same time; periodic vibrations induce resonant modes at integer multiples of the fundamental frequency ) . Here , we examine whether a statistical inference model can learn these cues directly from natural sounds; and if so , how effective are these learned cues relative to existing hand-tailored segregation systems . The proposed model is designed as a hierarchical system that explicitly mimics an ‘analysis-then-fusion’ processing pipeline . The analysis stage is itself laid out in two stages . First , an analysis of local spectrotemporal cues aims to learn simultaneous Gestalt cues believed to operate over short-time scales in order to locally segregate sound elements . Second , an analysis of more global cues operates over longer time-scales and aims to learn sequential Gestalt cues that enable tracking dynamics of elements from the first stage at a temporal or melodic level [8] . Following these stages is a fusion step that combines together segregated elements that constitute different auditory objects , using principles of temporal coherence [39 , 64 , 65] . The Gestalt analysis stages are learned directly from natural sounds in a generative fashion , allowing each component of the model to represent natural sounds from its own vantage point following principles of stochastic neural networks , as detailed next . The fusion stage merely organizes or fuses these learned patterns following the concept of temporal coherence , as also detailed later . Fig 1 depicts a schematic of the overall model . It takes as input the acoustic waveform of an auditory scene u ( t ) and maps it onto a time-frequency representation , using a biomimetic peripheral model from Yang et al . [66] . Briefly , this transformation analyzes the acoustic signal u ( t ) using a bank of logarithmically-spaced cochlear filters whose outputs are further sharpened via a first order derivative along the frequency axis , followed by half wave rectification and short term integration over 10ms frames ( see Methods for details ) . This filterbank analysis results in an auditory spectrogram represented by S ( t , f ) . The following stage ( called L 1 ) is structured as a two-layer sparse Restricted Boltzmann Machine ( sparse RBM ) with a fully connected visible and hidden layer [67] . It takes as input 3 consecutive frames of the spectrogram and learns a probability distribution over the set of these short tokens . RBMs are powerful stochastic neural networks that are conceptually similar to autoencoders but can infer statistical distributions over their input set [68] . A RBM layer is chosen for this stage in order to explore the space of local spectrotemporal tokens and learn latent cues that represent statistical structures in natural sounds over short time scales . The visible layer units {xk} are real-valued and characterized by a Gaussian distribution fitted over the input spectrogram S ( t , f ) ; while hidden units {hk} are sampled from a Bernoulli distribution for k = 1 , 2 , … , K where K is the number of nodes in each layer . The network is parameterized by Θ = {W , A , B} where W represents the interconnected weights between visible and hidden units , and A ( B ) represents the visible ( hidden ) bias , respectively . The network is trained using a Contrastive Divergence ( CD ) algorithm with the objective to minimize the reconstruction error between x and x ^ = h W + A [69] . By learning the regularities in local spectrotemporal tokens of natural sounds , the connection weights W effectively span an array of latent cues that reflect the structure of soundscapes . Our hypothesis is that these latent factors represent the so-called simultaneous cues used as Gestalt principles for sound analysis . After training , connection weights are transformed into a 2D filter F ( t , f ) , akin to spectro-temporal receptive fields derived from neural activity of biological neurons in the auditory system [70] . These learned filters are then applied in a convolutional fashion over the incoming spectrogram S ( t , f ) to derive the outputs of layer L 1 nodes . These responses are further subjected to a neural adaptation stage which imposes a dynamic regulation of the response of each filter hence suppressing units with weak activation ( see Methods for details ) . L 1 responses are then processed by the next layer in the model which completes the analysis stage to infer possible sequential cues that extend over longer time constants . This second layer L 2 is devised as an array of conditional RBMs ( cRBMs ) , which are extended versions of RBMs designed to model temporal dependencies [71] . Similar to a RBM , a cRBM consists of a visible layer with units {xk} , assumed to arise from a Gaussian distribution fitted over the input , and a hidden layer with {hk} units sampled from a Bernoulli distribution . Unlike a RBM , a cRBM acts as a dynamical system operating over an entire input history τ taking as input occurrences at times {t , t − 1 , … , t − τ} in order to capture dynamics in the input space over context τ . In the current model , we explore sequential cues over a range of temporal contexts and construct an array of parallel cRBM networks over multiple histories ranging in temporal resolutions from τ ∼ ( 30–600 ms ) . L 2 is parameterized by Θ = {W , Aτ , Bτ , Cτ , Dτ} where W represents the interconnected weights between visible and hidden units and capture the interactions across input features over an extended temporal history τ , Aτ and Bτ represent the visible and hidden biases , respectively , while Cτ and Dτ quantify autoregressive weights between past inputs and the current input ( or current hidden unit , respectively ) . Just like the localized layer L 1 , the contextual layer L 2 is trained in a generative fashion using contrastive divergence ( CD ) in order to best capture the dynamics in natural sounds using the same dataset of realistic sounds spanning speech , music and natural sounds . Here again , our hypothesis is that the stochastic cRBM learns latent parameters Θ that reflect the sequential cues underlying dynamics of natural sounds over a wide range of temporal contexts . Once trained , the model parameters are applied to incoming L 1 filter responses in a linear fashion , yielding a multi-resolution output which is then passed over to the next stage in the hierarchy ( see Methods for details ) . The next layer in the hierarchy focuses on a fusion operation to facilitate the grouping of perceptually-coherent objects . This binding stage explores co-activations across all L 2 channels within a given context τ and binds together the units that exhibit strong temporal coherence [64 , 72] . The ‘temporal coherence’ theory posits that emergence of perceptual representations of auditory objects depends upon strong coherence across cues emanating from same object and weaker co-activation across cues from competing objects . This coherence is not an instantaneous correlation but one that is accumulated over longer time scales , commensurate with the contextual windows explored in the L 2 layer . We implement this concept in a biologically-plausible fashion via mechanisms of Hebbian learning , which suggests that when two neurons fire together , their synaptic connection gets stronger [73] . Effectively , Hebbian interactions operate by reinforcing activity across coherent channels , hence grouping them into putative objects and inhibiting activity across incoherent channels [74] . We implement a synaptic interaction across output channels from layer L 2 by introducing a coherence synaptic weight matrix V . If two units i and j are co-activated at a given time t , their corresponding synaptic connection Vij is reinforced over time . If the correlation between their activity is weak , the corresponding synaptic weight Vij is reduced accordingly . These synaptic weights are applied to the output of each channel in a dynamic fashion , hence modulating the activity across an entire ensemble of neurons within each context in layer L 2 . The net effect gives emergence to perceptual coherent groups that represent auditory objects in a scene . A final read-out stage is then appended to the model to extract responses to different stimuli and test the degree of segregation of different objects , as viewed by the model outputs ( see Methods for details ) . In order to examine the emergent sensitivity of learned layers in the network , we derive the tuning characteristics of individual nodes or neurons and explore their filtering properties in the modulation domain [75 , 76] . Modulation tuning reflects stimulus cues that best drive individual nodes in the model both in terms of temporal variations and dynamics ( i . e . temporal modulations or rates ) as well as spectral span and bandwidth ( i . e . spectral modulations or scales ) . This approach follows common empirical techniques used in electrophyisology and psychophysics to probe the tuning of a system to specific acoustic cues . It is specifically used to characterize spectro-temporal receptive fields ( STRFs ) which offer 2-dimensional profiles of filtering characteristics of neurons [70] . First , we employ a classic transfer function method using probe stimuli in order to derive the tuning of both L 1 and L 2 layers of the network [77–79] . We present modulated noise signals ( called ripples ) as input to the model with varying spectro-temporal modulation parameters ( Fig 2E ) and characterize the fidelity of the ripple encoding at various stages of the network as the ripple modulation parameters are varied [80] . Each ripple is constructed as a broadband noise signal whose envelope is modulated both in time and frequency , with temporal modulation parameter ω ( in Hz ) and spectral modulation parameter Ω ( in cyc/oct ) ( see Methods for details ) . By sweeping through a range of ripple parameters , we compute a normalized modulation transfer function ( MTF ) from the response of layers L 1 and L 2 which quantifies the synchronized response of each layer to the corresponding dynamics in the ripple stimulus ( see Methods for details ) . L 3 is not a trained layer and hence is not subject to this analysis . Fig 2A and 2B depict the MTF derived from both L 1 and L 2 . The functions highlight that both layers exhibit a general low-pass behavior both along temporal and spectral modulations . As expected , layer L 1 is trained over shorter time-scales and does exhibit faster temporal dynamics along the rate axis , while the contextual layer L 2 is mostly tuned to slower dynamics < 30Hz with a slightly tighter spectral selectivity mostly concentrated below 1 cycles/oct . This outcome is very reminiscent of similar transfer functions obtained from neurophysiological data showing contrasting tuning characterizations in the midbrain , auditory thalamus and auditory cortex [81–83] , whereby selectivity of individual neurons along the mammalian auditory hierarchy evolves from faster to slower temporal dynamics and from more refined to broader spectral spans along frequency . We further examine the selectivity of individual neurons and compare emergent tuning characteristics common across nodes in the network by employing an agglomerative clustering algorithm ( see Methods for details ) . This approach clusters nodes exhibiting similar tuning profiles into common groups hence providing insight into the underlying acoustic cues being processed by each cluster . Fig 2C and 2D show contour plots from the resulting clusters overlaid on the MTF profiles for layers L 1 and L 2 . The array of clusters indicates that neurons in each of these layers do indeed exhibit a wide variety of selectivity to spectral and temporal dynamics in the input signal . We specifically note a cluster of L 1 neurons that is more sensitive to fast transients or ‘onsets’ . This group is labeled ‘O’ in Fig 2C . An example time-frequency profile F ( t , f ) of a neuron in the ‘O’ cluster is shown in Fig 2F ( upper-right ) . We also note a spectrally-structured cluster ( labeled ‘H’ ) centered around spectral modulations ∈ [1-2] cyc/oct corresponding to harmonic peaks present in natural sounds . An example neuron from this cluster is shown in Fig 2F ( upper-left ) and highlights the selectivity to specific frequency bands in the input spectrogram . The clustering procedure also reveals the presence of oriented spectro-temporally selective clusters , likely tuned to detect frequency-modulated sweeps in the signal over different spectrotemporal scales; as well as other clusters with special selectivity to spectral or temporal features . Fig 2F ( lower panels ) shows an example of two L 1 neurons with different temporal dynamics contrasting a slow neuron ‘S’ and a fast neuron ‘F’ . We test the model’s behavior with a variety of acoustic scenes ranging from classic streaming paradigms using simple tones to experiments using speech signals . Crucially , all experiments are tested on the same model ( after all layers have been trained ) , without any adjustment to model parameters . The stimulus parameters are carefully chosen to closely replicate previously published human perceptual experiments hence allowing a direct comparison between the model and human perception . All stream segregation results are shown in Fig 3 organized in 3 columns: the stimulus on the left , a replica of human perception of the same stimulus reproduced from the corresponding publication in the center , and the model performance on the right . As outlined earlier , Fig 3 contrasts the model’s performance against reported human perceptual results in a range of stream segregation experiments . Next , we reexamine our initial hypotheses; namely that the model is able to infer simultaneous and sequential grouping cues by learning statistical regularities in natural soundscapes . The experimental results shown in the previous section suggest that simultaneous cues ( tonotopic organization , AM rate , harmonicity , temporal synchrony , etc ) , sequential cues and grouping mechanisms play an important role in streaming paradigms . In order to shed light on their individual contributions , we run a series of control experiments where we look at malfunctions in the model if certain components of the system are disrupted individually .
The analysis of control experiments quantifies the complementarity of rich feature representation and grouping mechanisms in driving scene segregation . The proposed architecture faithfully replicates human psychoacoustic behavior on steaming paradigms over wide range of stimuli ranging from simple tones to speech utterances as demonstrated in Fig 3 . In case of two tone streaming paradigm shown in ( Fig 3A ) , the network exhibits stream segregation when two alternating tones are widely separated across tonotopic frequency axis . This behavior is consistent with well established psychophysical and physiological findings of stream segregation induced by differences in tonotopic cues [129–132]; and relies heavily on the activation of different groups of neurons with distinct frequency selectivities as captured in L 1 . In absence of temporal correlation between these two groups , the temporal coherence layer aided by the adaptation mechanism suppresses the anti-correlated groups of units , hence inducing stream segregation in the final stage of the network . However when ΔF is small enough , there is high degree of overlap resulting in a single stream percept . This segregation/integration effect is strongly maintained regardless of a number of manipulations to the model architecture . The key components crucial to the organization of tone sequences are the presence of tonotopic or frequency selectivity combined with temporal integration that examines activity across neural channels at relatively longer time-scales . This observation is very much in line with the spatio-temporal view of auditory stream segregation which requires neural channels to be widely separated in addition to temporal asynchrony across these channels [133] . The interaction of spectral and temporal dynamics during the organization of tone sequences supports the view of stream segregation as a dynamic process . The buildup effect reported in the current model ( Fig 3B ) is in line with established psychoacoustic behaviors [90 , 134–136] and suggests that segregation of two streams is not instantaneous; but strengthens over time and can lead to segregation when frequency difference ( ΔF ) is large enough . The current model highlights that this effect is in fact reflecting the competition across neural channels as viewed by the temporal coherence layer . The binding of correlated groups of neurons strengthens over time while suppressing the anti-correlated units over time in the same process . Interaction across multiple features is also noted in other simulations that pit against each other harmonicity , onsets and temporal dynamics ( Fig 3[C] , 3[D] and 3[E] ) . Simulations using complex tones directly examine the role of localized spectro-temporal tuning in L 1 as an encoding of simultaneous cues such as harmonicity , onset and fast amplitude modulations among others . Sequential cues emergent in L 2 are crucial in tracking the activity emerging in the localized layer over longer amplitude modulations; which are then fused together in the last L 3 layer . Through this rich selectivity learned directly from natural sounds , the network offers a wide span of selectivity across the spectrotemporal space . This tuning proves effective in tackling complex auditory scenes composed of speech with various interferers . In line with human perceptual data , the model shows that speech tokens are harder to identify in presence of utterances from same corpus compared to babble and cafe noise as the signal-to-noise ratio gets smaller . The model highlights that this variable response is largely caused by the dominance of neural activity from the interfering set relative to the target . The distinct activation between target and interferer is further blurred in absence of of slow sequential cues which integrate information about the speech utterance beyond just that target number/color . As shown in the control experiments , a network that lacks slow sequential cues is further impaired in making a judgment about the identity of the target token , likely due a to an enhanced confusion between its representation and that of the interferer . Once this activity reaches the temporal coherence layer , the weakly responsive neurons get suppressed , hence resulting in the actual number/color token getting wrongly identified as the one in the interfering utterance . Overall , the proposed model highlights three key results: ( i ) Using the right configuration , we are able to infer a wide-range of Gestalt cues directly from natural sounds . The proposed RBM architecture offers a cooperative and nonlinear integration of these cues to result in a multiplexed representation of auditory scenes across various granularities in time and frequency . By using an unsupervised learning approach , the network is not being optimized for a specific application; rather , it is reflecting the inherent variety of local and global dynamics present in natural sounds . Possibly , an even deeper neural architecture extending beyond just a few layers could extend the rich feature analysis and fill in the spectrum from local to global hence adding a more refined mapping along with the nonlinear integration naturally offered by the RBM architecture . ( ii ) Grouping acoustic features is effectively an outlook across all active nodes that allows to piece together the pieces of each auditory object . This process effectively plays 2 key roles: a grouping role by putting together pieces of a sound object ( effectively integrating together pitch , timbre , rhythm and possibly space information that reflect a common object ) ; and an elimination role by suppressing channels that are irrelevant to the emergence of the foreground object , hence enhancing the signal-to-noise ratio in the network . Temporal coherence is one such fusion mechanism that has been garnering stronger neural and perceptual evidence [39 , 65 , 125 , 126] . The current work employs Hebbian learning , a biological simple mechanism that affords such fusion over the rightly chosen time-scales . ( iii ) Auditory scene segregation is a balancing act of the proper feature analysis along with mechanisms for fusion that give rise of auditory object representations . While both stages are necessary , neither one is sufficient . The proposed model offers a unified platform that integrates together these different mechanisms and strategies . It also bridges the existing physiological theories of scene organization with perceptual accounts of auditory scene analysis .
The proposed model is structured along 4 key stages: initial data pre-processing by transforming the acoustic signal to a time-frequency representation , a local analysis over short time-scales , a global analysis over an array of longer time-scales , then a fusion stage using temporal coherence . A final readout of the network activity is implemented to extract information from specific streaming experiments to probe segregation of individual streams in the input scene . Details of each component of the model are outlined next: The acoustic signal is first analyzed through a model of peripheral processing in the mammalian auditory system , following the model by Yang et al . [66] . Briefly , it transforms the acoustic stimulus sampled at 8KHz into a joint time-frequency representation referred to as auditory spectrogram . The stage starts with a bank of 128 asymmetric constant-Q filters equally-spaced on a logarithmic axis over 5 . 3 octaves spanning the range 180 Hz to 4000 Hz ( QERB ≈ 4 ) [137] . By its very nature , the peripheral model uses a non-parametric set of cochlear filters that are fixed over a span of 5 . 3 octaves ( see [66] for details ) . In the current model , we cap our sampling rate to 8KHz in order to provide ample coverage over lower frequency regions . After cochlear filtering , the outputs undergo spectral sharpening via first order derivative along frequency , followed by half-wave rectification then short term integration with e−t/τ where τ = 10 ms . This filterbank analysis results in a time-frequency auditory spectrogram represented by S ( t , f ) . Three consecutive frames are then grouped together to form a one dimensional vector x such that x ∈ Rn and n = 384 . This process is repeated for all the audio samples in the dataset to form a set of N sampled patches given by X = x1 , x2 , … , xN . This set of time-frequency patches ( X ) constitutes the input to second component of the network . We test the model on stream segregation paradigms spanning tones , complexes and speech; as detailed next . For all non-speech simulations , the final readout compares the model response to a given stimulus and to a slight variation of that stimulus in order to probe whether their respective outputs exhibit noticeable differences , which would indicate a segregated or grouped percept . Ultimately , the model readout quantifies the response difference between these signals ( as a relative measure ) as we sweep through the input parameters . This approach is consistent with classic techniques used to objectively probe stream segregation in human listeners ( see [86] for more discussion ) . In the present study , a threshold is chosen empirically to quantify the difference between the stimulus and its variant in order to label it as 1 stream ( small enough difference ) or 2 streams ( large enough difference ) . In all cases , we confirm that the results are qualitatively similar when we vary the choice of thresholds within a reasonable range . Details of this comparison procedure are outlined below . The procedure for segregation of speech signals is different , as specified in the speech intelligibility section . Two tone sequences . In order to determine whether the tones in the ABA tone sequence are grouped into a single stream or multiple streams , we alter the last burst of the A tone by 4% of its actual frequency in either direction ( upward or downward ) in one sequence ( represented by A’ ) and keep the A tone the same in another sequence . We pass both sequences through the model and compute the Euclidean distance between final responses obtained for the sequence with change and sequence with no change . As the separation between A and B tones increases , we notice that this Euclidean distance increases . We determine an empirically chosen threshold over this distance measure to indicate whether tones A and B are grouped into a single stream or form segregated streams . A d’ measure is then used to quantify correct ( hit rate ) and false detection ( false alarm ) of A for both alternating and synchronous sequence; which is computed as: d ′ = z ( H ) − z ( F ) ( 15 ) where z ( ) represents the z-score . The d’ score determines the strength of auditory streaming , in line with the approach used in the psychophysical results reported in [86] . Buildup effect on stream segregation . The analysis of buildup also alters the the final burst of the sequence as either tone A or A’ as explained earlier . If the network can report any difference between A and A’ based on a thresholded Euclidean distance , we consider A as a single stream , otherwise both A and B are grouped into single stream . Here , we used the percentage of correct detection of tone A as metric to determine streaming , consistent with results reported in [90] . Amplitude-modulated noise sequences . In the noise sequences , the final burst is comprised of either noise A having the modulation rate of 100 Hz or noise A’ with slight alteration of 10% to actual modulation rate . The results are then reported in terms of percentage of correct detection of noise A following a similar thresholded Euclidean measure approach . Tone complexes with harmonicity and onset variations . Just like previous experiments , the final burst of the sequence in each trial comprise of either target tone A or an alteration of 4% to the actual frequency of A in random direction represented by A’ . A d’ analysis based on correct ( hit rate ) and false detection ( false alarm ) of A for all possible combinations is reported following the procedure described earlier . Speech intelligibility . The model’s performance is assessed based on a simplified speech identification task that only employs a readout of the encoding of the target speech segments in the model . First , we divide all utterances belonging to a particular target ( either number or color ) into training and test sets . Each of the utterances is passed through the entire network to obtain an output response . Frames belonging to the target token are collected together and their corresponding output responses are averaged out to get a single mean response for each utterance . We collect all such responses across the entire training set and build GMM models [145] for each target . The test utterance is then passed trough the network to obtain the corresponding output response and averaged across the frames corresponding to the target token similar to training paradigm . This average response is then analyzed through each of the GMM models to obtain the log likelihood score relative to each target P ( target|θ ) where θ represents the GMM parameters for each target class . Based on a predetermined threshold defined empirically , a decision is made as to whether the system identifies the correct target token or not . We repeat the experiments for all the colors and numbers in the CRM corpus and report the accuracy of the system in terms of percentage correct identification of color , number and both color and number . | In every day life , our brain is able to effortlessly make sense of the cacophony of sounds that constantly enter our ears and organize them into meaningful sound objects . In this work , we use an architecture based on stochastic neural networks to ‘learn’ from natural sounds which cues are crucial to the process of auditory scene organization . The computational model delivers a hierarchical architecture that mimics multistage processing in the biological auditory system . It learns a rich hierarchy of spectral and temporal features that allow the decomposition of an auditory scene into informative components . These features are then grouped together into coherent objects based on Hebbian learning principles . Though trained on unrelated datasets of natural sounds , the model is able to replicate human perception of auditory scenes in a wide variety of soundscapes ranging from simple tone sequences to complex speech-in-noise scenes . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"acoustics",
"linguistics",
"neural",
"networks",
"engineering",
"and",
"technology",
"signal",
"processing",
"social",
"sciences",
"neuroscience",
"neuronal",
"tuning",
"computer",
"and",
"information",
"sciences",
"animal",
"cells",
"speech",
"physics",
"cellular",
"neuroscience",
"speech",
"signal",
"processing",
"cell",
"biology",
"neurons",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"physical",
"sciences",
"cognitive",
"science",
"modulation",
"acoustic",
"signals"
] | 2019 | A Gestalt inference model for auditory scene segregation |
In eukaryotic cells , ribosomal RNAs ( rRNAs ) are transcribed , processed , and assembled with ribosomal proteins in the nucleolus . Regulatory mechanisms of rRNA gene ( rDNA ) transcription and processing remain elusive in plants , especially their connection to nucleolar organization . We performed an in silico screen for essential genes of unknown function in Arabidopsis thaliana and identified Thallo ( THAL ) encoding a SAS10/C1D family protein . THAL disruption caused enlarged nucleoli in arrested embryos , aberrant processing of precursor rRNAs at the 5’ External Transcribed Spacer , and repression of the major rDNA variant ( VAR1 ) . THAL overexpression lines showed de-repression of VAR1 and overall reversed effects on rRNA processing sites . Strikingly , THAL overexpression also induced formation of multiple nucleoli per nucleus phenotypic of mutants of heterochromatin factors . THAL physically associated with histone chaperone Nucleolin 1 ( NUC1 ) , histone-binding NUC2 , and histone demethylase Jumonji 14 ( JMJ14 ) in bimolecular fluorescence complementation assay , suggesting that it participates in chromatin regulation . Furthermore , investigation of truncated THAL proteins revealed that the SAS10 C-terminal domain is likely important for its function in chromatin configuration . THAL also interacted with putative Small Subunit processome components , including previously unreported Arabidopsis homologue of yeast M Phase Phosphoprotein 10 ( MPP10 ) . Our results uncovering the dual role of THAL in transcription and processing events critical for proper rRNA biogenesis and nucleolar organization during reproduction are the first to define the function of SAS10/C1D family members in plants .
The biogenesis of mature 5 . 8S , 18S , and 25S ribosomal RNAs ( rRNAs ) requires transcription of 45S rRNA genes ( rDNA ) and processing of 45S precursor rRNAs ( pre-rRNAs ) in the nucleolus [1] . The nucleolus is not enclosed by a membrane; its formation is driven by the active transcription of rDNA and structured by pre-rRNA processing and ribosome assembly components . rDNA units are tandemly arrayed at nucleolar organizer regions ( NORs ) , and NORs of Arabidopsis thaliana ( Arabidopsis ) abut upon the northern telomeres of chromosomes 2 and 4 ( NOR2 and NOR4 , [2] ) . The four NORs present in a diploid cell collectively form a single nucleolus , with active rDNA decondensed inside the nucleolus where they undergo transcription and silenced rDNA in compact heterochromatin blocks at the external periphery of the nucleolus [3] . Silent rDNA units are densely methylated at their promoters and associated with modifications such as histone 3 lysine 9 methylation ( H3K9me ) ; active rDNA are hypomethylated and enriched with H3K4 trimethylation ( H3K4me3 ) [3 , 4] . Currently , the rRNA regulatory network underlying the structure and function of the nucleolus remains evasive , and machinery components involved are yet to be defined . One major component in the nucleolus is the Small Subunit ( SSU ) processome , a ribonucleoprotein ( RNP ) complex required for biogenesis of 18S rRNA and subsequent assembly and maturation of the ribosome SSU in yeast Saccharomyces cerevisiae [5] . It contains the U3 small nucleolar RNA ( snoRNA ) and U Three Proteins ( UTPs ) , with a total of as many as 72 non-ribosomal proteins which compose numerous subcomplexes [6] . A subset of SSU processome components , called t-UTPs for their requirement for transcription , are necessary for optimal rDNA transcription and closely associated with ribosomal chromatin [7] . Therefore , rDNA transcription and pre-rRNA processing are functionally connected , but to date there are limited reports investigating the coupling and co-regulation of these two processes [8] . Although extensively studied in yeast , the SSU processome is not validated in many other organisms including Arabidopsis . The Something About Silencing 10 ( SAS10 ) /C1D family proteins contain the SAS10/C1D and/or SAS10 C-terminal domains . In yeast and mammals , members of this family were shown to participate in RNA processing , translational control , DNA repair , and gene silencing [9] . For instance , yeast rRNA Processing 47 ( RRP47 ) is an exosome cofactor required for processing of rRNAs and snoRNAs [10] . RRP47 interacts with exosome catalytic subunit RRP6 via its SAS10/C1D domain [11] . Its mammalian homologue C1D functions as a DNA repair factor by interacting with and activating the catalytic subunit of the sensor of DNA double-strand breaks , DNA-Dependent Protein Kinase ( DNA-PK , [12] ) . Both RRP47 and C1D binds RNA as well as DNA , and it was proposed that SAS10/C1D domain simultaneously serves as a platform for protein interactions and a nucleic acid binding site [9] . On the other hand , SAS10 C-terminal domain has not been formally investigated . Currently there are no published reports of any members of SAS10/C1D family in plants . Here , we present the characterization of a member of plant SAS10/C1D family named Thallo ( THAL ) . thal-2 arrested embryos contained enlarged nucleoli likely caused by over-accumulated pre-rRNAs; THAL overexpression gave rise to multiple nucleoli which may be the result of ectopic transcription and dispersal of rDNA . The interacting partners of THAL include Nucleolin 1 ( NUC1 ) , Arabidopsis homologue of yeast M Phase Phosphoprotein 10 ( AtMPP10 ) , and Nucleolar Factor 1 ( NOF1 ) of the putative SSU processome in Arabidopsis , and possibly NUC2 and H3K4me2/3 demethylase Jumonji 14 ( JMJ14 ) . Combining these findings , we propose that THAL contributes to both transcription and processing pathways of rRNA biogenesis and thereby impacts the organization of the nucleolus and reproductive development .
An in silico forward genetic screen for Arabidopsis transfer DNA ( T-DNA ) insertional mutants with only non-homozygous progeny from the SALK Homozygote T-DNA collection identified several mutants defective in gametophytic or embryonic development . One of these mutants harbored a T-DNA insertion in the seventh intron of At2g43650 ( Fig 1A ) . The encoded protein comprises two putative domains: SAS10/C1D and SAS10 C-terminal domains characteristic of SAS10/C1D family members ( Fig 1A ) . This protein shares low ( 22% ) identity with yeast SAS10 ( S1A Fig ) . Phylogenetic analysis shows the widespread presence of its orthologues in other higher eukaryotes ( S1B Fig ) , suggesting essential functions of these proteins . The expression of At2g43650 was detected by reverse transcription PCR ( RT-PCR ) in all tested tissues , including shoots , rosette and cauline leaves , flowers , siliques , roots , and seedlings , with highest expression in shoots and flowers ( S2A Fig ) . We thus named this gene Thallo ( THAL ) , after the Greek goddess of buds and shoots . Detailed spatial expression patterns were examined by β-glucuronidase ( GUS ) reporter assay , using transgenic plants expressing a GUS reporter gene under the control of a nearly 2-kb sequence upstream of THAL ( THALpro::GUS ) . Significant activity of the putative THAL promoter was observed in the subapical region of primary roots , lateral root primordia , leaf veins , and around guard cells in seedlings , as well as the ovule , pollen , embryo , and endosperm in adult plants ( S2B and S2C Fig ) . Collectively , THAL appears to be ubiquitously expressed , with preference for tissues undergoing rapid cellular growth and differentiation . In addition to the first mutant , designated thal-1 , another T-DNA line , thal-2 with the T-DNA inserted in the twelfth exon was acquired ( Fig 1A ) . Compared to wild type ( WT ) , heterozygous ( HZ ) plants containing either one of these two alleles did not show any apparent morphological abnormalities during all developmental stages except in developing siliques ( Fig 1B and Table 1 ) . thal-1/+ siliques had a portion of undeveloped seeds , resulting in an average of 37 seeds per silique , approximately three-fourths that of WT . thal-2/+ contained one-fourth of pale-yellow seeds in siliques 10 days after pollination . These defective seed development phenotypes could be complemented by the full-length THAL genomic fragment ( THAL/thal; Fig 1B and Table 1 ) . According to the Mendelian segregation ratio , we speculated that the abortive and yellow seeds in thal-1/+ and thal-2/+ siliques , respectively , represented homozygous ( HM ) progeny . Mature dry seeds from thal-1/+ and thal-2/+ were plated on half MS plates and one-fourth of seeds from individual thal-2/+ siliques failed to germinate ( S2 Table ) . Germinated seedlings from thal-1/+ and thal-2/+ had a 1:1 and 2:1 ratio of HZ to WT , respectively ( S1 and S2 Tables ) . Reciprocal crosses between thal-1/+ and WT showed that the transmission of thal-1 alleles is highly reduced through female gametes and slightly reduced through pollen ( Table 2 ) . Developing siliques of WT , thal-1/+ and thal-2/+ were cleared to visualize the embryonic stages of seeds ( Fig 1C ) . Though all seeds were presumably WT or HZ ( WT/HZ ) in thal-1/+ siliques , a minority of embryos were delayed in growth ( Fig 1C and 1D ) . In thal-2/+ siliques , approximately one-fourth of seeds arrested uniformly at globular stage whereas the remaining three-fourths had already developed to the torpedo stage ( Fig 1C and 1D ) . The arrested globular embryos ( embryo propers ) were often shaped in irregular spheres , suggesting defects in cellular division patterning ( Fig 1E ) . In addition , they were larger than WT globular embryos because they were older ( other embryos from the same silique had reached the torpedo stage ) . To ascertain whether the arrested globular embryos in thal-2/+ siliques were indeed due to disruption of THAL , we inspected T2 developing seeds from the thal-2 mutant complemented by an N-terminal GFP-tagged THAL coding sequence under THAL native promoter ( THALpro::GFP-THAL/thal-2 ) . Clear GFP signals were detected in globular and torpedo/bent cotyledon embryos in green seeds , but not in the irregular globular embryos found in yellow seeds ( S3A Fig ) . Given that WT embryos start to accumulate chlorophyll at the heart stage , the arrested globular embryos contributed to the pale-yellow appearance of seeds . These results strongly support the assumption that early embryo arrest is caused by loss of THAL function . The T-DNA is inserted towards the C-terminal end of THAL in thal-2 ( Fig 1A ) . To determine whether thal-2 expresses truncated THAL , we used total RNA extracted from green and pale-yellow immature seeds in thal-2/+ developing siliques ( hereafter bent cotyledon WT/HZ seeds and thal-2 seeds ) , as well as from seeds with globular-stage embryos in WT siliques ( globular WT seeds ) . Full-length THAL coding sequence was not detected by RT-PCR in thal-2 seeds , but truncated N- and C-terminal fragments of THAL were both detected in low levels ( S3B Fig ) . N-terminal fragment was produced probably because T-DNA is inserted at C-terminal end . C-terminal fragment may have been expressed by the promoter of immediate adjacent gene At2g43660 which is in a reversed orientation downstream of THAL . Hence , thal-2 is an embryo-lethal mutant that likely expresses truncated THAL and is used in following studies . To explore the possible molecular function of THAL , its subcellular localization was first examined by transiently expressing a C- or N-terminal GFP-tagged THAL coding sequence ( THAL-GFP and GFP-THAL ) under the CaMV 35S promoter in Arabidopsis protoplasts . The control GFP alone localized in the cytoplasm and nucleus of transformed protoplasts ( Fig 2A ) . Surprisingly , we observed recurrent multiple nucleoli in protoplast cells overexpressing THAL-GFP or GFP-THAL . The nucleolar area was defined by the absence of RFP-tagged nuclear marker Ethylene Responsive Transcription Factor 4 ( ERF4-RFP ) , which does not localize in nucleoli . Overexpressing another nucleolar protein , Fibrillarin 2 ( GFP-FIB2 ) , resulted in only a single nucleolus . Interestingly , GFP-THAL conferred a larger effect on nucleolar dispersion than THAL-GFP , as more than 90% of GFP-THAL—but less than 50% of THAL-GFP—overexpressed protoplasts contained multiple nucleoli ( Fig 2B ) . Above results were validated with transgenic plants harboring GFP-THAL driven by THAL promoter ( THALpro::GFP-THAL ) or the estradiol-inducible XVE chimeric activator ( XVEpro::GFP-THAL; S4 Fig ) . THALpro::GFP-THAL root cells exhibited GFP-THAL signals co-localizing with acridine orange ( AO ) nucleolar signals , thus confirming the nucleolar localization of THAL ( S4A Fig ) . In XVEpro::GFP-THAL plants , 20–50% of cells displayed multiple GFP-THAL foci indicative of multiple nucleoli after estradiol treatment ( S4B Fig ) . Taken together , the data demonstrate the nucleolar localization of THAL and that THAL overexpression induces formation of multiple nucleoli per nucleus . The GFP-THAL fusion protein was proven functional by a complementation experiment in which GFP-THAL but not THAL-GFP gave rise to viable HM plants . This result raised the possibility that a C-terminal GFP fusion perturbed the important function of SAS10 C-terminal domain of THAL . To address this hypothesis , we investigated the domains important for THAL localization by constructing a series of truncated THAL proteins fused to GFP at the N-terminus for protoplast transient expression assay ( Fig 2A , 2C and 2D ) . A C-terminal truncation lacking SAS10 C-terminal domain and the region between two domains ( GFP-THALΔC0 ) resulted in co-localization with ERF4-RFP in the nucleoplasm . GFP-THALΔC1 lacking only the SAS10 C-terminal domain localized in the nucleolus but did not induce formation of multiple nucleoli . By contrast , a C-terminal truncation lacking merely half of SAS10 C-terminal domain ( GFP-THALΔC2 ) could cause multiple nucleoli , albeit to a lesser extent than full-length GFP-THAL as most of these cells contained only two nucleoli . Finally , GFP-THALΔN with an N-terminal truncation lacking the N-terminal region and SAS10/C1D domain could no longer concentrate in the nucleolus and localized in nucleoplasm . Collectively , these results indicate that SAS10/C1D domain and the region between two domains are necessary for the nucleolar targeting of THAL , and SAS10 C-terminal domain is important for further regulation of nucleolar organization . It has been shown that translocated rDNA loci can retain their transcriptional activity and are able to self-assemble additional nucleoli [13] . We examined if the additional nucleoli observed in GFP-THAL ( and THAL-GFP ) overexpressed protoplasts are associated with NORs by fluorescence in situ hybridization ( FISH ) using 45S rDNA probes . The nucleolus contains mostly RNA and is not stained by the DNA-binding dye 4' , 6-diamidino-2- phenylindole ( DAPI ) . In Arabidopsis WT interphase cells , the 4 NORs tend to coalesce so usually 3 NOR signals are detected , 2 of which are associated with the nucleolus [14] . This is the case in GFP or GFP-FIB2 overexpressed cells ( Fig 3A and 3B ) . However , more than 4 NOR signals were discovered in protoplasts overexpressing GFP-THAL and they were mostly associated with nucleoli . There are four main rDNA variants ( VAR1-VAR4 ) in Arabidopsis Col-0 ecotype , based on insertions/deletions in the 3’ External Transcribed Spacer ( 3’ETS , [15] ) . Expression of each variant is differentially regulated among developmental stages and tissues . Given that multiple nucleoli observed upon THAL overexpression implies rDNA dispersal and transcription , the expression of individual rDNA variants were inspected by RT-PCR . The four variants were distinguished by a primer pair flanking the 3’ETS variable region ( Fig 3C ) . First we analyzed THAL loss-of-function effects , using total RNA extracted from globular WT seeds , bent cotyledon WT/HZ seeds and thal-2 seeds . Globular WT seeds showed similar expression levels of VAR1 , VAR2 , and VAR3 ( Fig 3D ) . Bent cotyledon WT/HZ seeds had highest expression of VAR3 . thal-2 seeds had similarly high levels of VAR2 and VAR3 . Therefore VAR1 , which represents nearly 50% of rDNA , was expressed in globular WT seeds but was inhibited in globular thal-2 seeds , suggesting that THAL is required for the activation of VAR1 . THAL gain-of-function effects were next analyzed using transgenic plants harboring CaMV 35S promoter driven GFP-THAL ( 35S::GFP-THAL ) . Interestingly , 35S::GFP-THAL plants did not exhibit obvious morphological defects before flowering but inflorescence stems failed to elongate in 7 out of 9 individual lines ( S5A Fig ) . Quantitative RT-PCR ( qRT-PCR ) analysis confirmed the overexpression of THAL in each line ( S5B Fig ) . It is noteworthy that THAL expression was highest in WT shoots ( S2A Fig ) . Hence , ectopic expression of THAL has adverse effects to plant development . We examined the expression of rDNA variants and found VAR1 is silenced in WT and 35S::GFP vegetative tissues , but is de-repressed in 35S::GFP-THAL plants ( Fig 3E ) . This is consistent with our previous results , which showed repressed VAR1 in thal-2 seeds . VAR2 and VAR3 expression also elevated in 35S::GFP-THAL as compared with 35S::GFP plants . We further examined the relative abundance of rDNA variants by PCR with genomic DNA and did not detect apparent differences in the proportions of rDNA variants among 35S::GFP-THAL , 35S::GFP , and WT ( S5C Fig ) . Thus , THAL is required for the activation of specific variants . Loss-of-function nucleolar phenotypes in thal-2 embryos were examined by Transmission Electron Microscopy ( TEM ) . TEM analysis of embryo sections showed that nucleoli , nuclei , and nucleolus to nucleus ratios were mostly larger in thal-2 embryos than those in WT/HZ bent cotyledon embryos from the same silique and those in WT globular embryos ( Fig 4A and 4B ) . Subnucleolar structures including fibrillar center , dense fibrillar component , granular component , and nucleolar vacuole did not show prominent differences in thal-2 ( S6 Fig ) . We generated thal-2/+ plants harboring a nucleolar marker FIB2-GFP by manual crossing and consistently found visibly larger nucleoli in thal-2 embryonic cells than those in WT/HZ bent cotyledon and WT globular embryos ( S7 Fig ) . The enlargement of nucleoli is phenotypic of pre-rRNA processing mutants in plants [16 , 17] . Pre-rRNA processing includes a series of cleavage events to remove the 5’ETS , 3’ETS , and internal transcribed spacers ( ITS1 and ITS2 ) for the generation of mature rRNAs . We examined the accumulation of pre-rRNAs in total RNA from globular WT seeds , bent cotyledon WT/HZ seeds , and thal-2 seeds . Due to the limited amount of materials , levels of pre-rRNAs were determined by qRT-PCR with two pairs of primers , one specifically amplifying a region in 5’ETS and another flanking a region in ITS1 ( Fig 4C ) . Additionally , three sets of primers amplifying the 18S , 5 . 8S , and 25S regions , respectively , were used to detect total rRNAs . Globular WT seeds had only half the fragments containing 5’ETS but similar levels of ITS1-containing pre-rRNAs as bent cotyledon WT/HZ seeds ( Fig 4D ) . However , thal-2 seeds accumulated approximately 2 . 5-fold more ITS1-containing pre-rRNAs and 5-fold more 5’ETS-containing pre-rRNAs than globular WT seeds . Total rRNA levels profoundly decreased in both thal-2 and globular WT seeds as compared with bent cotyledon WT/HZ seeds , but levels in thal-2 seeds were higher than those in globular WT seeds , which was likely due to the over-accumulation of pre-rRNAs ( Fig 4D and 4E ) . Thus , compared to bent cotyledon WT/HZ seeds , thal-2 seeds had over-accumulation of pre-rRNAs but probably much less mature rRNAs , thereby resulting in less total rRNAs; compared to globular WT seeds , thal-2 seeds had more pre-rRNAs and total rRNAs ( Fig 4E ) . The over-accumulation of pre-rRNA transcripts may have caused nucleoli in thal-2 embryos to become larger than those in WT globular embryos ( Fig 4A and 4B ) . Next , various processing sites ( cleavage sites ) were amplified in qRT-PCR , including P , P1 , P’ sites in 5’ETS , A2 , A3 , B1 sites in ITS1 , C2 site in ITS2 , and B0 site in 3’ETS . thal-2 seeds had a similar pattern but higher levels of all cleavage sites tested as compared with globular WT seeds when Elongation Factor 1α ( EF-1α ) was used as an internal control ( Fig 5A and 5B left panel ) , confirming the over-accumulation of pre-rRNA transcripts in thal-2 seeds . However , pre-rRNA over-accumulation could result from elevated transcription and/or impaired processing . In order to exclude transcriptional accumulation and assess only processing of pre-rRNAs , we also normalized to the full-length 45S precursor determined by a primer pair immediately after the transcription initiation site ( TIS; Fig 5A ) . Levels of nascent 45S precursors in thal-2 seeds were 4-fold higher than in globular WT seeds and similar to bent cotyledon WT/HZ seeds ( Fig 5C left panel ) . We found distinct accumulation patterns of cleavage sites between globular WT and bent cotyledon WT/HZ seeds , indicating that processing efficiencies differ among cleavage sites as well as embryonic stages ( Fig 5C right panel ) . In addition , P’ site amplification level was higher while P and P1 sites were lower in thal-2 seeds than in globular WT seeds . This suggests that THAL affects processing at the 5’ETS , of which it promotes P’ site cleavage but attenuates cleavage at P and P1 sites . Pre-rRNA processing in 35S::GFP-THAL primary inflorescence stems was next analyzed . Amplification of cleavage sites revealed a pattern overall opposite to that of thal-2 seeds , with P’ site amplification level becoming the lowest and P site level among the highest ( Fig 5B right panel ) . 35S::GFP-THAL accumulated more than 3-fold higher levels of 45S precursors than 35S::GFP ( Fig 5D left panel ) , which may be caused by VAR1 re-activation in 35S::GFP-THAL . There was over-accumulation of all cleavage sites compared with 35S::GFP even after normalization to 45S precursors ( Fig 5D right panel ) , likely because the increased amount of 45S precursors have overwhelmed processing components . Altogether , transcription was increased and processing events were delayed in 35S::GFP-THAL plants . While it is not clear how transcriptional enhancement would affect plant growth , delayed processing would further impair ribosome assembly and protein translation required for development . Northern blots were performed with 35S::GFP-THAL plants . Hybridization with S1 probe located in 5’ETS and downstream of P’ site detected the 35S ( P-B0 ) and P-A3 precursors in 35S::GFP as observed in WT of previous studies ( S10 Fig , [18 , 19 , 20] ) . However fragments larger than 35S and P-A3 were observed in 35S::GFP-THAL , suggesting that consistent with our qRT-PCR data , P site cleavage is indeed attenuated . S0 probe was used to further monitor the presence of pre-rRNAs containing the fragment upstream of P site , and results confirmed them over-accumulated in 35S::GFP-THAL . Finally , S2 probe situated in ITS1 upstream of A3 site detected the fragments recognized by S1 probe in addition to 32S and 18S-A3 fragments . In conclusion , these results confirmed that THAL is required for processing at the 5’ETS . In an effort to identify potential candidates for the interacting partners of THAL , we performed immunoprecipitation followed by mass spectrometry ( IP-MS ) analysis using total proteins extracted from 8-d-old rescued THALpro::GFP- THAL/thal-2 and WT seedlings . After eliminating those obtained in the WT sample , other than THAL itself , we identified several proteins involved in ribosome biogenesis: Ribosomal protein S6e ( RPS6B/EMB3010 ) , DEAD box RNA helicase RH3 ( EMB1138 ) , and Nucleolin 1 ( NUC1 ) in the GFP-THAL immunoprecipitate ( S11A Fig ) . Intriguingly , histone 2A proteins ( H2A . 1 , H2A . 2 , H2A . X . 3 , H2A . X . 5 , H2A . W . 6 , H2A . W . 7 , H2A . Z . 8 , H2A . Z . 9 , H2A . 10 , H2A . Z . 11 , and H2A . 13 ) were also detected . H2A . W . 12 was not detected probably because of its low abundance ( and HTA4 is a pseudogene ) . Among the associated proteins isolated by IP-MS , of particular interest was NUC1 , a multifunctional protein required for the expression of rDNA variants and NOR condensation [15] . NUC1 was also among the genes co-expressed with THAL ( S11C Fig ) . Additionally , the second nucleolin in Arabidopsis , NUC2 , and three proteins selected based on protein interaction prediction by the Bio-Analytic Resource database: Arabidopsis homologue of yeast M Phase Phosphoprotein 10 ( AtMPP10 ) , Jumonji 14 ( JMJ14 ) , and Nucleolar Factor 1 ( NOF1 ) were tested for interactions with THAL ( S11B Fig ) . Bimolecular fluorescence complementation assay ( BiFC ) was conducted using Arabidopsis protoplasts transformed with THAL and candidate protein fused to the N- and C-terminal fragments of YFP , respectively ( YFPN-THAL and NUC1-YFPC etc ) . Protoplasts co-expressing YFPN-THAL and NUC1-YFPC / NUC2-YFPC / AtMPP10-YFPC / JMJ14-YFPC / YFPC-NOF1 all yielded clear YFP fluorescence signals in nucleoli , while negative controls YFPN-THAL and YFPC empty vector or YFPN-NOT1 ( unrelated protein ) and NUC1-YFPC / NUC2-YFPC / AtMPP10-YFPC / JMJ14-YFPC / YFPC-NOF1 did not ( Fig 6A and S12 Fig ) . Co-localization of YFPN-THAL and NUC1-YFPC / NUC2-YFPC / AtMPP10-YFPC / YFPC-NOF1 ( all nucleolar proteins and used as nucleolar markers ) again demonstrates that the multiple foci are indeed nucleoli . We further verified THAL interactions by yeast two-hybrid ( Y2H ) assay , in which THAL consistently interacted with AtMPP10 and NOF1 ( Fig 6B ) . Interactions between THAL and NUC1 / NUC2 / JMJ14 were not detected , suggesting that their physical associations require a plant-specific component . Taken together , THAL associated with AtMPP10 , NOF1 , and NUC1 in multiple techniques ( BiFC and Y2H / IP-MS ) , but with NUC2 and JMJ14 only in BiFC . To further assay THAL interactions with NUC2 and JMJ14 , co-immunoprecipitation ( co-IP ) was performed using transient expression of GFP-THAL , NUC2-FLAG , and JMJ14-FLAG in Arabidopsis seedlings . NUC2-FLAG , but not JMJ14-FLAG , could be detected in the GFP-THAL immunoprecipitates ( Fig 6C ) . THAL and JMJ14 interaction could be specific to certain periods and conditions ( e . g . developmental stage and growth status ) or too transient to be detected . Likewise with NUC2 which merely showed a weak association . Nevertheless , THAL interaction with JMJ14 and NUC2 need to be further validated with caution .
Multiple nucleoli observed upon THAL overexpression resembles disrupted gene silencing . Previous studies have shown that interfering with essential heterochromatin regulators destabilizes nucleolar integrity [24–26] . In Drosophila melanogaster , H3K9 methylation and RNA interference pathways regulate the organization of rDNA and the nucleolus [24] . In Arabidopsis , telomerase-deficient cells displayed multiple nucleoli that occasionally coincided with extra rDNA signals [26] , a phenomenon we repeatedly observed in GFP-THAL overexpressed protoplasts . Chromatin decondensation increases recombination between DNA repeats , which results in dispersal of rDNA and ectopic nucleoli [24] . It is thus plausible that THAL plays a negative role in chromatin condensation and gene silencing , thereby affecting nucleolar integrity that requires NOR heterochromatic structures . Accordingly , THAL associated with histone chaperone NUC1 , histone-binding NUC2 , and H3K4me2/3 demethylase JMJ14 in our interaction experiments . Selective silencing of rDNA variants was just recently demonstrated to be chromosome-specific; rDNA variants located at NOR2 are silenced and those located at NOR4 are active [27] . In Col-0 vegetative tissues , VAR1 is located at NOR2 and silenced , whereas VAR1 introgressed into NOR4 genome is active . Therefore multiple nucleoli and de-repression of VAR1 upon THAL overexpression are likely results of dispersal and activation of NOR2 . Indeed , FISH results demonstrated more than 4 rDNA signals in GFP-THAL overexpressed protoplasts ( Fig 3A and 3B ) . Deduced negative role of THAL in chromatin condensation would further cause this dispersal of rDNA . There are two nucleolins found in Arabidopsis [28] . Disruption of NUC1 causes disorganized nucleoli and NOR decondensation . In nuc1 mutant leaves , VAR1 is de-repressed as in 35::GFP-THAL overexpression lines , thus THAL and NUC1 might play antagonistic roles in the regulation of VAR1 expression . Alternatively , THAL and NUC1 work in concert in a complex and THAL overproduction causes a shortage of NUC1 proteins , therefore mimicking nuc1 . A second nucleolin in Arabidopsis , NUC2 , expressed only during specific developmental stages , acts antagonistically with NUC1 [14] . NUC2 , along with JMJ14 , AtMPP10 and NOF1 , were not identified in our IP-MS experiment since we used 8-d-old seedlings in which NUC2 protein level is undetectable [14] and THAL interactions with these proteins might be stage-specific or transient . Differing from nuc1 but similar to THAL overexpression , increased rDNA loci association with the nucleolus was observed in nuc2 . However VAR1 was also de-repressed in nuc2 seedlings . Our results demonstrating THAL interacts with NUC1 and possibly NUC2 insinuate that THAL may regulate NUC1 and NUC2 functions . Further studies are anticipated to decipher the molecular mechanisms among these proteins during various developmental stages . NUC1 binds to activated VAR1 and conversely NUC2 binds to silent chromatin [14 , 15] . We hypothesize that THAL interacts with NUC1 ( or NUC2 , possibly depending on the developmental stage ) to contribute to transcription of rDNA variants ( Fig 7B ) . Furthermore , THAL might associate with NUC1 / NUC2 and JMJ14 at the histones to assist chromatin remodeling . However , since 45S precursors were accurately processed in nuc1 , we deduce that THAL , AtMPP10 , and NOF1 but not NUC1 / NUC2 in a presumed SSU processome participate in pre-rRNA processing at the 5’ETS . THAL is primarily expressed in differentiating and dividing cells where protein synthesis is a high demand ( S2B and S2C Fig ) . In thal , impaired production of mature rRNAs diminished subsequent ribosome assembly and concomitant protein translation , which terminally brought about early developmental arrest . Similar to THAL , nucleolins are multifunctional proteins implicated in various aspects of ribosome biogenesis , and nuc1nuc2 double mutant is seedling lethal [14] . Nevertheless , pre-rRNA processing single mutants are often viable , as are gene silencing mutants [29–31] . Based on the severity of its mutant phenotypes , THAL seems to be more crucial for development than the nucleolins and a non-redundant regulator of rRNA biogenesis . It would be interesting to determine whether the dual role of THAL and its importance in nucleolar organization are evolutionarily conserved among species , especially mammals in which nucleolar enlargement is a common feature of cancer cells .
Arabidopsis thaliana seeds of Col-0 ecotype were used in this study . Two T-DNA insertion lines , SALK_016916 ( thal-1 ) and SALK_036872 ( thal-2 ) , were obtained from the Arabidopsis Biological Resource Center ( ABRC ) . After stratification at 4°C for 48 hr , seeds were germinated in soil or half Murashige and Skoog ( MS ) medium solidified with 0 . 7% agar ( pH 5 . 7 ) , then grown at 22°C under a 16-hr light/8-hr dark photoperiod . The full-length genomic DNA fragment ( ~6 . 5 kb ) of THAL and genomic fragment without the downstream putative terminator sequence ( ~5 . 5 kb ) were PCR amplified using genomic DNA extracted from seedlings . The coding sequence ( CDS ) of THAL was amplified with seedling cDNA as template . For complementation of thal mutants , 6 . 5- and 5 . 5-kb genomic fragments were cloned into pMDC99 and pMDC107 ( with GFP ) vectors , respectively , and transformed into thal/+ plants by the floral dip method . For GUS expression analysis , THAL native promoter fragment ( THALpro; ~1 . 8 kb ) was cloned into pCB308 vector and the construct was transformed into WT plants . For protoplast transient expression assay , THAL CDS and its truncated fragments were cloned into pGEM-T Easy vector carrying the CaMV 35S promoter and GFP sequence . To produce transgenic plants carrying GFP-THAL , GFP-THAL was isolated from the transient expression construct by treating with restriction enzymes and then cloned into pER8 ( under estradiol-inducible XVE promoter; XVEpro ) and pPZP221 vectors . THALpro was additionally cloned into GFP-THAL/pPZP221 . XVEpro::GFP-THAL/pER8 and THALpro::GFP-THAL/pPZP221 were transformed into WT and thal/+ plants , respectively . To observe embryonic stages , developing seeds in siliques from WT , thal-1/+ and thal-2/+ plants were cleared in a chloral hydrate solution ( chloral hydrate: distilled water: glycerol 8:3:1 , w/v/v ) for 30 min after fixation in ethanol: acetic acid 3:1 for 1 hr . Cleared seeds were observed under an Observer Z1 microscope ( Carl Zeiss ) . Fresh embryos were isolated from immature seeds with forceps under a microscope ( Nikkon , SMZ645 ) , then mounted in 5% glycerol for laser scanning confocal microscopy ( Carl Zeiss , LSM510 ) to record fluorescent images . For TEM analysis , fresh developing seeds were collected from siliques and stored in fixation buffer containing 2 . 5% gluteraldehyde and 4% paraformaldehyde in 0 . 1 M sodium phosphate buffer , pH 7 . 0 at 4°C before ultra-thin sectioning . Embryo sections were observed under a Philips CM 100 TEM Microscope ( Philips Research ) at 80 KV . Images were obtained with a Gatan Orius CCD camera . Protoplasts were isolated from 4-wk-old Arabidopsis WT leaves using fungal cellulase and macerozyme to remove cell walls [32] . DNA transfection was performed using the PEG-calcium solution , followed by 16-hr incubation at 24°C . As a nuclear marker , 35S::ERF4-RFP was co-transformed . Empty vector and 35S::GFP-FIB2 were used as controls . Transformed protoplasts were observed under a laser scanning confocal microscope ( Carl Zeiss , LSM510 ) . FISH was performed as described in [33] , using Arabidopsis 45S rDNA probes . Nuclei were stained with DAPI in antifade mounting medium ( Vectashield , Vector Laboratories ) . Globular WT , bent cotyledon WT/HZ and thal-2 seeds or 5-wk-old 35S::GFP-THAL and 35S::GFP leaves and primary inflorescence stems were collected , frozen in liquid nitrogen , and stored at -80°C before RNA extraction by the RNeasy Plant Mini Kit ( Qiagen ) and removal of DNA contamination by the TURBO DNA-free Kit ( Applied Biosystems ) . Single-strand cDNA was synthesized with a random primer in order to detect rRNA transcripts . Quantitative PCR was performed using the Power SYBR Green PCR Master Mix ( Applied Biosystems ) with primer pairs designed by the Primer Express Software ( S3 Table ) . Plasmid pairs were co-transformed into the yeast strain AH109 following the manufacturer’s instructions ( Clontech ) . Primary transformants were first selected on Leu/Trp dropout ( -LW ) SD medium and confirmed again by colony PCR before growing on His/Ade/Leu/Trp dropout ( -HALW ) medium . For transient expression , Agrobacterium cultures were grown overnight and resuspended in infiltration media ( 5% sucrose , 5 mM MES , 200 μm acetosyringone ) to 1 . 0 OD600 . Arabidopsis 7-d-old AvrPto seeedlings were vacuum infiltrated ( 24 hr after 10 μm dexamethasone application ) and collected 4 d after infiltration . For immunoprecipitation , nuclei were first extracted with nuclei isolation buffer ( 0 . 25 M sucrose , 15 mM PIPES pH 6 . 8 , 5 mM MgCl2 , 60 mM KCl , 15 mM NaCl , 1 mM CaCl2 , 0 . 9% Triton X-100 , 1 mM PMSF ) and resuspended in nuclei lysis buffer ( 50 mM HEPES pH 7 . 5 , 150 mM NaCl , 1 mM EDTA , 1% Triton X-100 , 0 . 1% DOC , 0 . 1% SDS , 1 mM PMSF , 1x Roche protease inhibitor cocktail ) with crosslinking reagent dithiobis ( succinimidyl propionate ) ( 1 mM DSP ) . To quench crosslinking 50 mM Tris pH 7 . 5 was applied . Extracted nuclear proteins were then incubated with equilibrated GFP-trap beads ( Chromotek ) at 4°C for 1 . 5 hr under gentle agitation , followed by 3 times of washing with wash buffer ( 50 mM Tris pH 7 . 5 , 150 mM NaCl ) . Western blots were performed using α-GFP ( Santa Cruz ) or α-FLAG antibodies ( Sigma ) . Sequence data referred in this article can be found in the Arabidopsis Genome Initiative or GenBank/EMBL databases under the following accession numbers: AT2G43650 ( THAL ) , AT1G48920 ( NUC1 ) , AT3G18610 ( NUC2 ) , AT5G66540 ( AtMPP10 ) , AT1G17690 ( NOF1 ) , AT4G20400 ( JMJ14 ) , AT4G25630 ( FIB2 ) , and AT2G37620 ( ACT1 ) . | The rRNA regulatory network underlying the structure and function of the plant nucleolus is largely unknown . We identified a previously uncharacterized SAS10/C1D family protein THAL as a novel component important for both rRNA gene expression and processing , which further impacts nucleolar structure . thal-2 mutant had enlarged nucleoli whereas THAL overexpressed cells displayed multiple nucleoli resembling heterochromatin decondensation . THAL associated with the well-known histone demethylase JMJ14 , which implies not only a role of THAL in chromatin regulation but also a role of JMJ14 in rDNA transcription . In addition , THAL interacted with putative Small Subunit ( SSU ) processome components including nucleolin and previously unreported Arabidopsis homologue of yeast MPP10 , thus our work provides initial evidence of the existence and partial constituents of the putative SSU processome in Arabidopsis . Our study adds a piece to the rRNA network puzzle , provides insight into SAS10/C1D family function and prompts investigation of THAL mammalian homologues , especially since nucleolar dysfunction is characteristic of many diseases including cancer . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"plant",
"anatomy",
"plant",
"growth",
"and",
"development",
"plant",
"embryo",
"anatomy",
"brassica",
"developmental",
"biology",
"plant",
"science",
"model",
"organisms",
"seedlings",
"cell",
"nucleus",
"embryos",
"plants",
"cellular",
"structures",
"and",
"organelles",
"nucleolus",
"cotyledons",
"(botany)",
"research",
"and",
"analysis",
"methods",
"arabidopsis",
"thaliana",
"embryology",
"plant",
"embryogenesis",
"plant",
"development",
"gene",
"expression",
"seeds",
"ribosomes",
"biochemistry",
"rna",
"plant",
"and",
"algal",
"models",
"ribosomal",
"rna",
"cell",
"biology",
"nucleic",
"acids",
"embryogenesis",
"genetics",
"biology",
"and",
"life",
"sciences",
"non-coding",
"rna",
"organisms",
"fruit",
"and",
"seed",
"anatomy"
] | 2016 | Dual Role of a SAS10/C1D Family Protein in Ribosomal RNA Gene Expression and Processing Is Essential for Reproduction in Arabidopsis thaliana |
Vector control is still our primary intervention for both prevention and mitigation of epidemics of many vector-borne diseases . Efficiently targeting control measures is important since control can involve substantial economic costs . Targeting is not always straightforward , as transmission of vector-borne diseases is affected by various types of host movement . Here we assess how taking daily commuting patterns into consideration can help improve vector control efforts . We examine three tropical urban centers ( San Juan , Recife , and Jakarta ) that have recently been exposed to Zika and/or dengue infections and consider whether the distribution of human populations and resulting commuting flows affects the optimal scale at which control interventions should be implemented . We developed a stochastic , spatial model and investigated four control scenarios . The scenarios differed in the spatial extent of their implementation and were: 1 ) a response at the level of an individual neighborhood; 2 ) a response targeted at a neighborhood in which infected humans were detected and the one with which it was most strongly connected by human movement; 3 ) a limited area-wide response where all neighborhoods within a certain radius of the focal area were included; and 4 ) a collective response where all participating neighborhoods implemented control . The relative effectiveness of the scenarios varied only slightly between different settings , with the number of infections averted over time increasing with the scale of implementation . This difference depended on the efficacy of control at the neighborhood level . At low levels of efficacy , the scenarios mirrored each other in infections averted . At high levels of efficacy , impact increased with the scale of the intervention . As a result , the choice between scenarios will not only be a function of the amount of effort decision-makers are willing to invest , but largely epend on the overall effectiveness of vector control approaches .
Infectious diseases continue to place a considerable burden on global human health and will likely continue to do so for the foreseeable future . Despite successes such as the eradication of smallpox and reductions in the prevalence of falciparum malaria , there are also many examples of pathogens with emerging or intensifying transmission or exhibiting geographic spread . Many interacting factors are involved in this dynamic , including changes in temperature and rainfall , increasing levels of urbanization , changes in land use , and an increase in commuting and travel distances [1] . These exacerbating factors lend themselves especially well to arboviruses transmitted by the anthropophilic mosquito Aedes ( Stegomyia ) aegypti , which thrives in ( sub- ) tropical urban settings and is often the primary vector of the dengue , chikungunya , Zika , and yellow fever viruses . Variation in dengue transmission risk , for instance , is thought to be driven largely by rainfall , temperature , urbanization and socioeconomic factors [2 , 3] . Part of what makes this mosquito such an important vector in urban settings is that it undergoes its life cycle in close proximity to humans . In the larval stage , it develops in water-holding containers in or around homes , while as an adult it almost exclusively feeds on humans and can find suitable resting spots and oviposition sites in and around the domicile . Thus , with urbanization projected to rise from 54% of the world population in 2014 to 66% by 2050 [4] , the amount of suitable habitat for this synanthropic vector will likely only continue to increase . Human behavior also plays a critical role in determining the outcome of ( re ) emerging epidemics [5 , 6] . One type of behavior that influences both the global spread of pathogens and local clustering of cases relates to human movement , both in the form of local commuting and travel or migration over longer distances . In the case of vector-borne diseases the dispersal of vectors also has to be considered in interaction with the movement across different scales by human hosts , which can shape local patterns of transmission intensity [7–9] and contribute to the heterogeneity in exposure levels typical of vector-borne disease transmission [10 , 11] . In general , the implications for disease spread become complex once contacts between individuals are clustered in some manner ( e . g . , for spatial , household or sociological reasons ) instead of assuming random or well-mixed populations [12] . In the case of vector-borne diseases , such heterogeneous patterns in exposure to infective bites can have implications for disease surveillance , estimates of risk , and control [13] . For instance , if transmission hot spots and vector population sources and sinks occur in a particular vector-borne disease system , these could potentially provide targets for control interventions [14] . Considering the movement of humans could therefore conceivably improve our ability to target both the areas of origin and the areas where onward transmission may occur [15] , or at least to help reduce costs associated with those efforts . In the case of dengue , Zika or chikungunya , such transmission clusters likely emerge from the synergy between the short distances travelled by adult Ae . aegypti [16 , 17] and the multiscale movement by humans , ranging from international travel , to movement between urban centers , to day-to-day travel activities , such as commuting or going to social gatherings . For instance in Iquitos , Peru , clusters of dengue tended to be observed on greater than a 100 m radius , therefore likely involving a combination of both mosquito and human movement [18] . Other studies have supported the notion that movements within urban areas shape transmission and risk of exposure . For instance , visiting areas of risk , rather than the location of the home , was identified as a key driver of exposure to dengue in Iquitos [19] . Such day-to-day movement and travel of humans is especially pertinent for viruses transmitted by Ae . aegypti due to its diurnal biting activity , making it more likely that the place of exposure to bites is disconnected from the home environment [20] . An expected result is that a typical infected human case would lead to a number of additional human infections well outside the focal area of their home , in which control would typically be applied [19] . Understanding how different human contact structures and mixing patterns based on human movement can affect outbreak prevention and control outcomes has recently been receiving increasing attention via the use of network models . For instance , the efficacy of contact tracing has been shown to depend both on whether individuals mix assortatively or disassortatively , as well as the rate of contacts [21] . Likewise , for pathogens spread through close contact , such as influenza , vaccination strategies can potentially be optimized based on knowledge of the contact network in an area , such that strategies that make use of various aspects of this network can perform better than random immunization [22] . However , the exact role played by human movement and its impact on control strategies may vary based on the distribution and density of human populations , as well as socioeconomics or culture . For instance , commuting flows within urban centers can differ drastically between cities , based on their size and distributions of residential and work places . The resulting differences in mobility patterns can likewise affect epidemic spread [23 , 24] . For Aedes-transmitted viruses , especially dengue , examples of successful use of vector control interventions to reduce exposure risk certainly exist , such as a successful larval source management and public education strategy employed for a significant period in Singapore [25] . However , application of vector control interventions often fails to prevent arboviral outbreaks , possibly due to inadequate control responses [26 , 27] . The scope and level of response , as well as how long the response is sustained , are all recognized as critical factors that determine the overall success of implementation . Improving implementation strategies is thus critical to improving Aedes-borne virus control programs [27] . Given the importance of human movement and the focal nature of transmission , it is tempting to consider control strategies that explicitly make use of knowledge of human mobility patterns . For instance , in a previous study , we found that the relative efficacies of area-wide versus more targeted control responses in limiting outbreak size depended on both the daily rate of commuting and the proportion of patches that implemented control interventions [28] . At the same time , while strategies at the household level ( e . g . , a form of contact tracing whereby households connected to positive cases are treated ) may be appealing , contact tracing in urban tropical centers in developing countries may be logistically infeasible [19] . Here , we explore a number of straightforward vector control implementation scenarios using a simplified spatial model of Zika transmission by Ae . aegypti . Despite their simplifications , patch-based models can nonetheless provide important insights ( due to their more tracteable nature ) into questions related to the role of commuting and transmission and optimal control ( e . g . , [24 , 29] ) . The simplifications we use here relate largely to the scale of movement and targeting of control within the model ( i . e . , we use a relatively large patch size , namely at the level of the neighborhood , within which we assume humans are well-mixed ) . This model is used to develop general insights into when and how human movement and the distribution of humans over the urban landscape should affect decisions regarding the scale of interventions . We explicitly do not intend for this model to be predictive or provide advice to local vector control decision-makers , and would recommend household- or individual-level models specifically parameterized to local conditions for that . Here , we explore the relative efficacy and effort involved in vector control implementation under the following scenarios: ( 1 ) focal ( individual neighborhood ) control implementation; ( 2 ) two types of area-wide control; and ( 3 ) control implementation in both an individual neighborhood and in the neighborhood most strongly connected to it via a priori knowledge of human movement patterns . We specifically investigate whether the derived strategic insights would be consistent among tropical urban centers of varying sizes , layouts , human distributions , and movement patterns , or whether strategies would have to be tailored to individual locations . To do so , we ran extensive sets of simulations for models coarsely capturing the human distributions and movement patterns of San Juan ( Puerto Rico ) , Recife ( Brazil ) , and Jakarta ( Indonesia ) —all locations where Ae . aegypti has or could potentially lead to Zika epidemics .
We developed a discrete , stochastic version of a spatial compartmental vector-borne pathogen transmission model , which was previously used to explore spatial aspects of mosquito control on a simplified grid-based landscape [28] . Here , in this metapopulation model individual patches represent neighborhoods or districts . Briefly , the model is suitable for microparasites where within-host dynamics can safely be ignored and infection status can thus be modelled as a population-level characteristic . We assume a pathogen with a single host species and a single vector species , which is a reasonable assumption for arboviruses such as dengue , chikungunya , or Zika viruses , transmitted by Aedes aegypti in many urban tropical areas [30] . The model parameterization is based on Zika virus ( S1 Table ) . The human ( host ) and mosquito ( vector ) populations are tracked by infection status and life stage . Within each patch ( neighborhood or district of a particular location , indicated by subscript k ) , the human population ( Nh , k ) consists of susceptible ( Sh , k ) , exposed or latent ( Eh , k ) , infectious ( Ih , k ) , and recovered or immune ( Rh , k ) hosts . The vector population ( Nv , k ) is made up of immature ( Lv , k ) , susceptible ( Sv , k ) , exposed ( Ev , k ) and infectious ( Iv , k ) mosquitoes . See the supplementary material for the equations that describe the transitions between these compartments . We base the spatial configuration of the model on three areas where dengue and/or Zika virus epidemics have been known to occur in recent years . The purpose is simply to have a diversity of spatial distributions of neighborhoods in tropical urban settings . The cities which informed our spatial modelling were San Juan ( Puerto Rico ) , Recife ( Brazil ) , and Jakarta ( Indonesia ) , though other differences between those cities ( e . g . , rainfall patterns or seasonality ) were not considered . A figure with directed graphs and information on sizes and relative distances between patches is provided ( Fig 1 ) . We thus also kept entomological details as simple as possible , and adjusted the density-dependent rate of larval mortality for each neighborhood such that the vector:host ratio was similar in all neighborhoods and across settings ( ca . 5 female mosquitoes per human ) . We used the lowest available level of administrative boundary for each city ( i . e . , neighborhoods or districts ) to inform a model of human movement , as well as the structuring of vector control ( i . e . , we assume the most intense form of targeting occurs at the scale of one patch or neighborhood and ignore the role of movement or heterogeneity within patches ) . The choice of neighborhoods as our smallest spatial unit was driven by two reasons: we wanted to cover areas that were large enough so that we could ignore the movement of mosquitoes , and to keep the model simulations from becoming computationally too demanding . It is likely true that within these neighborhoods human populations are not homogeneously distributed or exposed to mosquito bites . To capture such heterogeneity ( which would even occur within households ) , one would need an individual-based approach with much finer granularity . As we were interested in the broad question of whether the distribution of humans over the landscape and their movement might affect vector control strategies , and the patch size has little effect on that question , we use this simpler model to derive strategic insights ( but would suggest using individual-based models that account for local spatial and temporal distributions of vectors , vector control measures that are already in place , and other forms of local parameterization , for predictive purposes and policy recommendations ) . We assume mosquito dispersal is sufficiently limited between neighborhoods relative to the movement of hosts that we can ignore the impact of mosquito movement . The size of the human population was derived from gridded data from the WorldPop project ( www . worldpop . org ) [31 , 32] and summed over the districts . Human movement reflected commuting , such that a proportion of the population of each neighborhood would spend time at another patch each day . Specifically , each host has a specific home patch ( k ) , but can potentially be exposed to infective bites in other patches ( j ) with a probability of δ . The distribution of hosts that commute to other neighborhoods depends on both the distance to and population size of the neighborhoods , following a gravity-type model [33] . Fj , k=|j||k|+|j||k||j|dj , k2 ( 1 ) Where |j| and |k| are the respective number of humans residing in neighborhoods j and k , and dj , k represents the distance between these neighborhoods . Distances between neighborhoods were calculated by taking the centroid of each neighborhood and using the “distm” function of the geosphere package in R [34] . These probabilities ( of commuting from patch k to each other patch j ) are then normalized over all neighborhoods . For each patch , we then have a matrix W with probabilities of remaining in the home patch ( with probability wk , k = 1−δ , or to any other patch with probability wj , k=δFj , k* . The distribution of hosts per day over patches is based on draws from a multinomial distribution with these probabilities . We explored four distinct control scenarios where the presence of infected humans in a given patch triggered responses that varied from highly focal to collective and area-wide . Specifically , we had 1 ) a focal or individual-patch level response; 2 ) a targeted response , in which control occurs in the patch with infections as well as the patch with which it is most strongly connected; 3 ) a limited area-wide response , wherein control occurs in all neighborhoods within a certain radius ( 0 . 33 x the mean distance among all neighborhoods in that locality–a value chosen to lead to a scenario intermediate between scenarios 1 and 4 ) of the patch with infective humans; and 4 ) a collective , area-wide response , wherein all neighborhoods initiate control . We also obtained baseline estimates of the progression of the epidemic in the absence of mosquito control . After a threshold of 2 infected humans is detected in a participating patch ( as many infections will be asymptomatic [35] , we assume detection occurs only for a proportion of infections , based on draws from a binomial probability distribution with probability of success of 0 . 5 ) , control is initiated after 1–30 days ( randomly chosen for each simulation ) to reflect a time delay associated with factors such as laboratory processing of samples , the incubation period of the virus , and health-seeking behavior of symptomatic individuals . To focus on the impact of the spatial configuration and scale of control , we used the same control method in each scenario . Specifically , we assumed larval control ( e . g . , the application of pesticides to larval development sites ) was used , modifying the base rate of death of larvae as follows: μc=−log ( 1−θ100 ) +μ2 , where θ is the effectiveness of the control method or the percentage of immature mosquitoes in a given patch that are killed per day due to the insecticide . The level of induced mortality was chosen randomly from values ranging from 1–100% ( drawn from a uniform distribution ) . For each scenario in each of the three localities , we performed sets of 2000 simulations . Each simulation had a 50-day burn-in period to allow mosquito population sizes to reach stable values , after which a single infected human was introduced to a randomly chosen patch . The infection dynamics were then followed for a two-year period , to capture both the major epidemic peak after initial ( re ) emergence and a significant stretch of time afterwards to allow the system to stabilize beyond transient dynamics . To gain insight into the usefulness of various control strategies , we compare the number of infected and recovered humans at the end of the two-year period for each of the 4 intervention scenarios to the scenario without any control , based on simulations where the initial introduction spread . We also investigated how the dynamics of control depend on assumptions regarding the efficacy of the control effort and the location of neighborhoods participating in the control effort by performing additional sets of simulations where a randomly chosen 20% of neighborhoods were assumed to not participate . For these particular sets of simulations , we assumed efficacy of larval control per patch was 60% . As a measure of the level of investment or effort associated with each scenario , we kept track of the total number of persons covered ( number of inhabitants per patch x number of days that patch initiated control , summed over all patches ) . Although assessing the financial and economic costs associated with the different larval control strategies were beyond the scope of this analysis ( and we therefore likewise ignore complications such as economies of scale ) , we draw upon health economic methods to estimate the net benefits of each of the four control strategies , using the ‘no control’ scenario as our reference case . We calculated the net benefit of each simulation as the number of infections averted multiplied by an investment threshold value , which here represents the number of persons a city or control program is willing to cover in order to avert one infected case , minus the number of persons covered in the control scenario [36] . We then used the R package BCEA [37] to graph the probability ( i . e . , the proportion over all simulations in which a given scenario had the greatest net benefit ) that each scenario was most effective for a given amount of effort across a range of investment thresholds .
In general , all four control scenarios substantially limited the severity of the outbreak , with the collective response in particular having a greater impact . In terms of infections averted over the two-year period , compared to the scenario without larval control , the focal and targeted responses were consistently the least effective . The relative impact of the intermediate area-wide scenario depended on the urban center ( e . g . , its size , the number of neighborhoods and the distribution of humans across the locality ) in which the epidemic took place ( Fig 2 ) . In San Juan , the increase in effort associated with the two area-wide interventions resulted in a concomitant increase in efficacy . In Recife , the focal and targeted response were only slightly worse than the intermediate area-wide response in terms of impact , while the collective response was significantly more effective . In contrast , in Jakarta , the three limited responses were much closer to each other in terms of impact . An exploration of individual simulations via scatter plots of infections averted versus the total number of people covered by larval control over the two-year period reveals the complex relationship between investment and efficacy ( Fig 3 , top panel ) . There are two distinct patterns visible in these scatterplots: an area where the four strategies differ largely in the extent of the number of persons covered , but with no clear improvement in outcome; and an area where scenarios that cover more persons also result in a greater number of infections averted . A similar distinction is evident when looking at the number of infections averted per simulation in relation to the assumption of efficacy of larval control at the patch-level ( i . e . , the induced level of larval mortality per day in a patch that enacts larval control ) —at lower levels of efficacy there is no discernable difference in outcomes between the interventions , while at higher levels the more collective responses do achieve a greater impact ( Fig 3 , lower panels ) . A possible explanation for this phenomenon is as follows ( Fig 4 ) : when the efficacy of larval control is low ( e . g . , 40% ) , this intervention is unable to contain the spread of the infection and regardless of the response type , control is triggered everywhere . In other words , the focal response and collective response become essentially equivalent . This is in contrast to control at higher levels of efficacy , where we see that under a collective response , control ceases when transmission is interrupted . Under the focal control strategy , however , the virus maintains itself in the metapopulation triggering waves of control throughout the two-year period . We additionally performed simulations where we assumed that control would not be feasible in a proportion of randomly chosen and varied neighborhoods ( 20% ) . Based on this we can investigate what the relative impact is of including or excluding specific neighborhoods from the control responses . Our results suggest ( and support findings from a different model and scenarios [38] that the spatial configuration of control also impacts the effectiveness of the control response ( S1 Fig ) . This is likely due to the connectivity structures of the patches ( derived from a gravity model ) . In part , this also appears to relate to the population size of each specific neighborhood ( Fig 5 ) . However , this result was most evident for San Juan ( R2 = 0 . 67 ) and was weaker in both Recife ( R2 = 0 . 56 ) and particularly Jakarta ( R2 = 0 . 11 ) . In these larger environments , other factors ( e . g . , distribution of patches ) likely become more prominent . All three settings displayed similar relationships between the level of investment ( number of persons covered per infected case averted ) and the net benefits of control . ( Fig 6 ) . We analyzed these relationships separately for low ( <50% ) and high ( ≥50% ) levels of efficacy of larval control . There is a striking difference in which interventions provide the best value . At high levels of efficacy , despite having to cover all patches at once in response to a trigger , the collective response is most likely to provide the best outcome for a given level of effort . When control is less efficacious , the results are more variable between the sites . At low investment thresholds , the focal response provides the best outcome . In San Juan , the focal response is the most effective for any level of investment threshold , while in the other locations the radius , targeted , and collective approaches become more appealing at higher levels of investment .
An important question in vector-borne disease control is how to structure prevention and/or control interventions optimally in space and time . This is particularly relevant for Aedes-transmitted arboviruses such as Zika , dengue , and chikungunya , which tend to be highly focal both temporally and spatially [39 , 40] . The fact that a large proportion of infected humans ( at least in the case of dengue ) remains asymptomatic yet contributes to ongoing transmission only complicates effective focal test and control strategies for such pathogens [41] . Given the tremendous burden on human health associated with Aedes-transmitted viruses , and the fact that the majority of this burden is experienced in tropical , often resource-constrained settings , there is a clear need for insight into what would constitute the most effective control strategy for a given amount of investment of effort . An example of this tension is provided by Liebman et al [18] , who suggest that while the current World Health Organization ( WHO ) guidelines recommend implementing vector control in a 400 m radius around a detected human case [42] , control responses in Iquitos , Peru tended to be limited to a radius of 100 m , as a 400 m radius could involve treating hundreds of houses , overly straining time and resources . Yet even the 400 m boundary suggested for perifocal responses would not limit transmission if responses are not immediate or the case in question has travelled outside the targeted area [42] . In other words , control responses to outbreaks could be vastly improved by taking the extent of movement into account . Here , we explored this topic using coarse models of human populations and their movement modelled after three different urban tropical centers by investigating the relative impact of different control implementation strategies that ranged from focal to collective , area-wide approaches . Our results illustrate that , as might be expected , collective ( city-wide ) responses are considerably more effective at limiting total outbreak size than are the more focal strategies . In general , the number of infections averted ( compared to a control scenario without any intervention ) increased with increasingly ambitious strategies . However , the relative impact of the two intermediate strategies ( treatment based on a radius around the focal neighborhood , and treatment in both the focal and the most strongly connected neighborhood ) differed between the three cities . The targeted approach in particular appeared to offer no benefit over the focal response . The reasons for these differences are not immediately obvious but could be tied to differences in scale ( population , physical size , number of neighborhoods , etc . ) . This suggests that optimizing control responses will have to be done on a location-by-location specific manner . Critically , however , when we estimated the effort involved in enacting each of the control scenarios ( by summing the total number of persons covered over all days of treatment ) and used this metric to investigate which scenarios were the most effective for a given amount of effort , the outcomes are remarkably consistent between the urban centers . When efficacy of control is high , the collective response provides the best value . This is likely because although it is costly to treat all locations simultaneously , this strategy then can succeed at quickly interrupting transmission of the virus . Two important caveats are that we only introduce the pathogen at a single time point , and that this high level of efficacy of larval control is unlikely to be achieved in reality due to the cryptic and abundant larval habitats employed by Ae . aegypti . For lower levels of efficacy , the scenario that offers the best value varies by investment threshold and location . At lower investment thresholds , the focal scenario is likeliest to provide the best value , while at high levels , the targeted response provides the best value in Recife and Jakarta , while in San Juan the focal response is consistently the most effective for a given level of effort . Whether such insights would still be evident were we to extend this modelling approach with appropriate costing and economic models was beyond the scope of the current study , but would be required before making policy recommendations . We also investigated whether there might be differences in the most effective spatial configuration of control , and discovered that there were certain neighborhoods which , if included in the control scenario , tended to increase or decrease the overall effectiveness . The neighborhoods whose inclusion most increased effectiveness tended to have lower human population sizes in San Juan , though this relation was somewhat ( in Recife ) and considerably ( Jakarta ) weaker in the other two environments . Because we used a constant ratio of vectors to hosts , this outcome does not relate to differences in vectorial capacity directly . As our movement was informed by a gravity model , smaller neighborhoods will receive fewer commuters . Larger neighborhoods , on the other hand , will attract visitors from a larger number of patches and therefore be better mixed , which can lower the intensity of transmission . This would suggest that in certain areas it may be worthwhile to target less well-mixed neighborhoods . It is not clear why the impact of neighborhood size on effect size decreased in the larger cities . It is possible that above a certain scale , factors related to the spatial distribution or configuration of patch sizes becomes more relevant . The criteria for selecting specific neighborhoods for control may in that case have to take the scale of urban centers into account . There were a number of simplifying assumptions of the model and simulations . For instance , human movement as simulated here is coarse and captures only commuting behavior at a between-neighborhood level , rather than fine-grained distinctions that in reality occur between the distances travelled , time spent at different locations , and frequency at which particular types of habitats ( e . g . , residential , recreational , commercial , etc . ) are visited [43 , 44] . Similarly , we assumed that the movement patterns could be well-described with a gravity model , and for some settings , other models of movement may capture reality better ( e . g . , radiation models [45] ) . In other words , the current study represents an important first step at investigating how the level of coordination between different neighborhoods may vary for different spatial compositions of humans , but it is certainly not an exhaustive exploration of the implications of human movement patterns for disease control . One particularly interesting complication relates to the effect of sickness behavior of febrile cases on mobility patterns . In the case of dengue , it has been qualitatively shown that individuals presenting with fever and testing positive for dengue visited fewer locations and spent more time inside their home than afebrile study participants [46] . It is likely that such effects will vary over the course of infection and depend on the severity ( or lack ) of symptoms . Including such sickness-mediated changes in behavior in a transmission and control model as used here was beyond the scope of the current study , but would be important to consider as both the transmission dynamics and implications for control strategies could change drastically . Another assumption we made relates to the likelihood of infections triggering control . We based this on two humans , and assumed the probability of detecting these infections was informed by an estimate of an symptomatic:asymptomatic ratio of 1:1 ( e . g . , [35] ) . This may still be an overestimate of the probability of detecting an infection , either due to health-seeking behavior or inefficiencies in health care systems . We additionally assumed control would only be triggered upon detection of two infected humans , whereas a single case may be used in reality ( e . g . , [47] ) . These respective over- and underestimates of triggering events will cancel each other out to an extent . The implications of varying the trigger for Zika control has been investigated recently [48] . Although costs are not captured here in an economic or financial sense , we do provide a crude measure of effort by looking at the total number of persons that would be covered by control efforts . This , therefore , disallows comparisons on a monetary scale , and also necessarily ignores a variety of complications ( e . g . , future discounting , ( dis ) economies of scale [49] ) that are all necessary to consider in actual cost-effectiveness studies . Thus , although our approach provides general insight into the relative efficacy of the various strategies , it would be necessary to consider actual costs associated with the various scenarios in future studies . To conclude , we have investigated both how the spatial scale of control strategies and the choice of neighborhoods which are included in interventions may influence the efficacy of control in limiting outbreak size . Our results suggest that the overall best choice of strategy will largely depend on decision-makers’ willingness and ability to invest in control measures ( which is likely to be location specific and may even vary within cities , depending on the organization of vector control ) , and the efficacy of control methods . The spatial configuration or distribution of control ( i . e . , the neighborhoods whose inclusion in scenarios were likely to increase the effectiveness of the scenario ) appeared to differ by urban center , where for San Juan ( a relatively smaller city ) and to a somewhat smaller extent Recife , the inclusion of neighborhoods was at least in part driven by the human population size , whereas targeting based on this factor became less important in Jakarta . While this finding needs further investigation , it adds to the conclusion that there will be no one-size-fits-all optimal solution for vector control strategies between different environments . Further work is needed to determine how such strategies should be structured , and to what extent that will vary between different urban centers of different sizes , distributions , and movement patterns of humans , under different frequencies of introductions of pathogens , and tailored to individual environmental , seasonal , and entomological settings . | Control and prevention of Aedes-transmitted viruses , such as dengue , chikungunya , or Zika relies heavily on vector control approaches . Given the effort and cost involved in implementation of vector control , targeting of control measures is highly desirable . However , it is unclear to what extent the effectiveness of highly focal and reactive control measures depends on the commuting and movement patterns of humans . To investigate this question , we developed a model and four control scenarios that ranged from highly focal to area-wide larval control . The distribution of humans and their commuting patterns were modelled after three major tropical urban centers , San Juan , Recife , and Jakarta . We show that as implementation is applied across a wider area , a greater number of infections is averted . Critically , this only occurs if the efficacy of control at the neighborhood level is sufficiently high . A consistent outcome across the three settings was that the focal strategy was most likely to provide the best outcome at lower levels of effort , and when the efficacy of control was low . These outcomes suggest that optimal control strategies will likely have to be tailored to individual settings by decision makers and would benefit from localized cost-effectiveness modelling studies . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"dengue",
"virus",
"medicine",
"and",
"health",
"sciences",
"pathology",
"and",
"laboratory",
"medicine",
"togaviruses",
"pathogens",
"neighborhoods",
"vector-borne",
"diseases",
"microbiology",
"social",
"sciences",
"human",
"mobility",
"alphaviruses",
"viruses",
"chikungunya",
"virus",
"rna",
"viruses",
"infectious",
"disease",
"control",
"human",
"geography",
"infectious",
"diseases",
"geography",
"medical",
"microbiology",
"microbial",
"pathogens",
"urban",
"areas",
"flaviviruses",
"earth",
"sciences",
"geographic",
"areas",
"viral",
"pathogens",
"biology",
"and",
"life",
"sciences",
"organisms",
"zika",
"virus"
] | 2019 | Contrasting the value of targeted versus area-wide mosquito control scenarios to limit arbovirus transmission with human mobility patterns based on different tropical urban population centers |
Gastrointestinal involvement affects 30–40% of the patients with chronic Chagas disease . Esophageal symptoms appear once the structural damage is established . Little is known about the usefulness of high resolution manometry to early identification of esophageal involvement . We performed a cross-sectional study at the Vall d’Hebron University Hospital ( Barcelona , Spain ) between May 2011 and April 2012 . Consecutive patients diagnosed with Chagas disease in the chronic phase were offered to participate . All patients underwent a structured questionnaire about digestive symptoms , a barium esophagogram ( Rezende classification ) and an esophageal high resolution manometry ( HRM ) . A control group of patients with heartburn who underwent an esophageal HRM in our hospital was selected . 62 out of 73 patients that were included in the study fulfilled the study protocol . The median age of the Chagas disease group ( CG ) was 37 ( IQR 32–45 ) years , and 42 ( 67 . 7% ) patients were female . Twenty-seven ( 43 . 5% ) patients had esophageal symptoms , heartburn being the most frequent . Esophagogram was abnormal in 5 ( 8 . 77% ) . The esophageal HRM in the CG showed a pathological motility pattern in 14 patients ( 22 . 6% ) . All of them had minor disorders of the peristalsis ( 13 with ineffective esophageal motility and 1 with fragmented peristalsis ) . Hypotonic lower esophageal sphincter was found more frequently in the CG than in the control group ( 21% vs 3 . 3%; p<0 . 01 ) . Upper esophageal sphincter was hypertonic in 22 ( 35 . 5% ) and hypotonic in 1 patient . When comparing specific manometric parameters or patterns in the CG according to the presence of symptoms or esophagogram no statistically significant association were seen , except for distal latency . The esophageal involvement measured by HRM in patients with chronic Chagas disease in our cohort is 22 . 6% . All the patients with esophageal alterations had minor disorders of the peristalsis . Symptoms and esophagogram results did not correlate with the HRM results .
Chagas disease ( CD ) is caused by the hemoflagellate protozoan , Trypanosoma cruzi . CD is endemic in Latin America . However migratory flows have spread the disease all over the world; the non-endemic countries with more cases are The United States of America and Spain . [1] Its prevalence worldwide is around 10 million cases and 25 million people remain at risk . [2] Patients with the chronic form of CD usually remain asymptomatic for years . Decades after the infection around 30–40% of the patients will develop visceral involvement caused by direct lesion , inflammation and fibrosis of the affected organs . The disease affects mainly the heart and the gastrointestinal tract , leading in some cases to dilated cardiomyopathy , megacolon and megaesophagus . [3] Gastrointestinal involvement ranges between 5–35% of patients , showing lower frequencies in studies from non-endemic countries , despite being similar populations . [3 , 4] Dilatation of the digestive tract and motor disorders , such as achalasia , delayed gastric emptying and altered colonic transit , are found in affected patients . [5] The pathogenesis behind the digestive involvement lays on enteric nervous system injury . [6] Dysphagia is usually the first symptom of esophageal involvement in CD , and achalasia and dilatation of the esophagus the main findings . [7] Contrary to what is seen in idiopathic achalasia , the pressure in the lower esophageal sphincter ( LES ) in patients with CD is diminished , [8] reflecting an impairment in both excitatory and inhibitory innervations . Unfortunately , the treatment for esophageal involvement in CD is addressed to symptoms relief . Classically , esophageal assessment in CD is performed with barium swallow . Conventional manometry has also been used in research studies . Recently , new tools that provide a greater knowledge about the esophageal involvement such as high resolution manometry ( HRM ) are available; however data of their use in CD are still scant . The objective of this study is to describe the findings of the esophageal HRM in patients with CD and to correlate esophageal symptoms with parameters and patterns of the esophageal HRM .
Consecutive CD patients attending the Tropical Medicine Unit of the Vall d’Hebron University Hospital ( Barcelona , Spain ) between May 2011 and April 2012 were invited to participate . Patients were referred to our units from blood donors centers , primary care centers , mother-to-child transmission prevention program and emergency service or hospitalization . Some patients came directly to other unit for a voluntary screening . Inclusion criteria included a minimum age of 18 years old and confirmed diagnosis of CD in the chronic phase . Patients previously treated for Chagas disease , pregnant women and patients who declined to sign the informed consent were excluded . All subjects underwent ECG and chest X-ray . Treatment with benznidazole was offered to all patients after the esophageal assessment . A control group was selected from patients with heartburn who underwent an esophageal HRM in our Hospital from April 2009 to April 2012 . This control group was selected because heartburn was the most frequent symptom among the CD cohort . Patients in the control group were born in Spain ( except for one patient that was from Equatorial Guinea ) and had no risk factor for CD . All patients in the control group were surveyed using the same questionnaire as the CD group . Diagnosis of CD was based on two positive serological enzyme-linked immunosorbent assay ( ELISA ) tests , one with recombinant antigen ( Bioelisa Chagas , Biokit , Lliçà d’Amunt , Spain ) and the other with crude antigen ( Ortho T . cruzi ELISA , Johnson & Johnson , High Wycombe , United Kingdom ) . Additionally , T . cruzi DNA was assessed in the peripheral blood by qualitative PCR before initiating treatment with benznidazole . PCR assay was performed according to Piron et al[9] . Rome III criteria were used to evaluate digestive symptoms through a personal interview . The interviewer had a structured questionnaire to guide the process . [10] The following items were included in the questionnaire: heartburn , dysphagia to liquids and solids , chest pain and regurgitation . All patients underwent barium esophagogram and high resolution manometry . Results of the esophagograms were classified according to Rezende classification . [11] All the manometric studies were carried out with HRM , consisting in a solid-state catheter with 36 circumferential sensors , ( each of them having 12 different entry ports ) , spaced by 1-cm distance along the intracorporal part of the catheter assembly ( Sierra Scientific Instruments Inc . , Los Angeles , CA , USA ) . A fasting of a minimum of 8 hours was required before the procedure . All the drugs that could interfere with the esophageal motility were discontinued . The catheter was introduced transnasally with the patient seated , until the most distally recording sensors were correctly placed in the stomach . Once positioned , the catheter was fixed in place by taping it to the nose . Subsequently , the patient adopted the supine position and the beginning of the protocol was postponed 1 minute approximately , to facilitate the relaxation and stabilization of the esophageal motility . The protocol started with the measurements of the basal sphincters’ pressure after a 30-second period . During this time , the patient was requested to breathe normally and not swallow . A minimum of 10 swallows of water of 5-ml each were then administered spaced by 30 seconds . As a supplemental test , a final multiple rapid swallowing was performed in every patient previously to the catheter removal . Manometric data were analyzed using dedicated software , ( ManoView , Given Imaging , Yoqneam , Israel ) , following the indications reported by Pandolfino et al . [12] The esophagogastric junction ( EGJ ) was firstly characterized . The proximal and distal limits were identified as abrupt increased changes in the pressure relative to both the intraesophageal and the intragastric pressures . Once identified , the integrated relaxation pressure ( IRP ) was recorded . An IRP equal or lower than 15 mmHg was considered normal . The distal esophageal body contractions were analyzed by the generation of isobaric contour plots at 30 mmHg . The mean values of the variables: Basal upper esophageal sphincter ( UES ) pressure ( normal value 34–104 mmHg ) , relaxing UES pressure ( normal value <12 mmHg ) , contractile front velocity ( CFV ) ( normal value <9 . 0 cm/s ) , distal contractile integrated ( DCI ) ( normal range 450–8000 mmHg s cm ) and distal latency ( DL ) ( normal value >4 . 5 seconds ) were recorded in each study . [13] The overall motility pattern was firstly classified following the 2012 Chicago classification criteria and reviewed after the recent publication of the new guidelines . [14 , 15] Thereby , the main categories considered were disorders with EGJ outflow obstruction ( achalasia types I , II and III and EGJ outflow obstruction ) , major disorders of peristalsis ( distal esophageal spasm , jackhammer esophagus , absent contractility ) , minor disorders of peristalsis ( ineffective motility and fragmented peristalsis ) and normal esophageal motility . To avoid observer-dependent bias , all the HRM reports were analyzed by two experimented gastroenterologists who were unaware of the symptoms of the patient . If any disagreement was encountered , it was brought to discussion and the final decision was decided in consensus . Data were analyzed with IBM SPSS Statistics software ( v . 21 . 0 . 0 . 0; IBM SPSS , Armonk , NY ) . The median and interquartile range ( IQR ) were calculated for quantitative variables . Frequencies and percentages were calculated for qualitative variables . Analysis was performed using Student’s t-test or Mann–Whitney’s U test for quantitative variables and Chi-square test or Fisher’s test for qualitative variables when appropriate . Tests were considered significant when the two-tailed p-value was <0 . 05 . The study protocol was approved by the Ethical Review Board of Vall d’Hebron Hospital ( Barcelona , Spain ) and procedures were carried out in accordance with the ethical standards laid down in the Helsinki Declaration as revised in 2000 . All patients signed the informed consent after a careful explanation of the study .
Initially , 73 patients with a confirmed diagnosis of CD were included in the study . Sixty two patients fulfilled the study protocol and were analyzed . The remaining 11 patients did not come to perform the HRM and declined to reschedule the test . Thirty patients were included in the control group . In the CD group , the median age was 37 ( IQR 32–45 ) years . Forty two ( 67 . 7% ) patients were female . All the patients in the CD group were migrants from Latin American countries ( 60 ( 96 . 8% ) patients were from Bolivia , one from Paraguay and one from Equador ) . When comparing baseline characteristic between the two groups , the control group was older ( 37 vs 51 years; p<0 . 001 ) and had more esophageal symptomatology ( 43 . 5% vs 93 . 3%; p<0 . 001 ) than the CD group . More information is shown in Table 1 . The clinical questionnaire was carried out in all included patients . Almost half of the CD group ( 27 patients ( 43 . 5% ) ) was symptomatic . From the symptomatic patients with CD , none of them referred a disabling intensity in any of the symptoms . Overall , the most frequent symptom was heartburn , which was referred by 19 ( 30 . 6% ) patients . This was largely followed by regurgitation and dysphagia to solids . Among patients in the control group , 28 ( 93 . 3% ) patients had at least one symptom and 6 ( 20% ) patients referred disabling symptoms . Clinical data are summarized in Table 1 . The esophagograms were performed in 57 patients ( 92% of the CD group ) being normal in the vast majority of the cases ( 52 patients ( 91 . 23% ) ) . Only 5 ( 8 . 77% ) patients had hiatus hernia . The esophagogram assessment was not performed in the control group . Fig 1 shows examples of HRM tracings . A pathological motility pattern was seen in 14 ( 22 . 6% ) patients in CD group . The HRM pattern more frequently found was minor disorders of the peristalsis ( 13 with ineffective esophageal motility and 1 with fragmented peristalsis ) . Hypotonic LES was present in 13 patients in the CD group , while in the control group 1 patient had hypotonic LES and 3 patients had hypertonic LES ( p = 0 . 005 ) . UES was normal in 39 ( 62 . 9% ) patients , hypertonic in 22 ( 35 . 5% ) and hypotonic in 1 patient in the CD group . In the control group 18 ( 60% ) patients had a normal UES , 9 ( 30% ) a hypertonic UES and 3 a hypotnic UES . More data can be found in Table 2 ( HRM technical parameters ) and in Table 3 ( manometric patterns according to Chicago classification ) . When comparing manometric parameters or patterns in the CD group according to any esophageal symptom no statistically significant association was seen , except for the median values in the DL parameter ( 7 . 2 symptomatic patients vs 6 . 5 non-symptomatic patients; p = 0 . 049 ) , however all the patients were considered to have normal values . Data are depicted in Table 4 .
Our study describes the main esophageal HRM findings in a cohort of consecutive patients with chronic CD . It reports the presence of low-intensity esophageal symptoms in almost half of the patients in the CD group , but there was no association between esophageal symptoms and the findings in the esophageal HRM . In CD the esophageal involvement is mainly characterized by the loss of the myenteric plexus . This translates in a motility pattern similar to that described in idiopathic achalasia , consisting in simultaneous contractions of low amplitude in the esophageal body , however LES in CD patients with esophageal involvement tends to be normal or hypotonic . [16 , 17] These alterations have been thoroughly studied with conventional esophageal manometry , however , few studies to date have reported the main pathological findings using HRM . Herein , no specific inclusion criteria regarding the symptoms were applied to avoid selection bias , and therefore , obtain a representative sample of the chronic CD population . Thus , 56 . 5% of the cohort was completely asymptomatic . This percentage is consequent with current literature findings , reporting values that range between 33–52% . [18 , 19] These percentages cannot be interpreted as a lack of esophageal involvement due to CD as it has been described that symptoms do not arise until the neuronal damage is extensive , [20] and as our study shows the vast majority of the patients had a normal esophageal HRM despite low esophageal symptoms . The first symptom usually developed in the course of esophageal involvement in patients with CD is dysphagia , however depending on the study the results may vary . [7] In our study , the most frequent symptom was heartburn , while dysphagia was present only in few patients . Heartburn in patients with CD is frequently accompanied by normal esophageal tests , while dysphagia goes normally with altered esophageal tests . Heartburn is a very common symptom in the community and its positive predictive value for esophageal involvement in patients with CD is low . This observation may be explained by the low esophageal involvement in our cohort and goes accordingly with previous studies . [18 , 21] When great damage is already established in the esophagus , heartburn gives way to dysphagia as the main symptom . On the other hand , it should be mentioned that 10 patients of our study referred dysphagia ( 3 of them with dysphagia to both solid and liquids ) . In 9 of these cases , the barium esophagogram was strictly normal , ( one patient did not have this test performed ) , and in 8 patients the HRM was informed as completely normal . The remaining 2 had minor disorders of the peristalsis which may be considered an early manifestation of CD esophageal involvement . Hence , it seems possible that dysphagia in these patients was not a consequence of CD as this symptom appears once there is a large damage of the esophagus . The results of the diagnostic tests were also remarkable for the absence of significant pathological findings . None of the barium esophagograms showed any grade of dilatation , with the only abnormality being a hiatus hernia . Based on our findings , we suggest that barium esophagograms should not be performed as part of the initial evaluation of every patient with chronic CD , due to its lack of discrimination of early esophageal alterations . HRM provides clear advantages with respect of conventional manometry in the characterization of the esophageal contractions and , more importantly , in the definition and features of the LES . [22] As a result , it has been postulated that the application of this technique in the evaluation of chronic CD would lead into the description of motility patterns that may have been unnoticed with conventional manometry . Surprisingly , in our cohort the motility alterations were remarkably mild . For instance , 77 . 4% of the sample had a complete normal HRM report . In the remaining patients , 14 patients had minor disorders of the peristalsis . When comparing with the heartburn control group , the basal pressure of the LES was lower in the CD group and the rate of hypotonic LES was also statistically lower . These findings are in concordance with current knowledge that states that the involvement of the esophagus of CD is due to destruction of the enteric nervous system , affecting both inhibitory and excitatory nerves [6] . There are four studies that assessed esophageal HRM in patients with chronic CD . All of them were performed in endemic countries . Two studies performed by Vicentine et al [23 , 24] in Brazil included patients with CD and dilatation of the esophagus and one study by Silva et al [25] also performed in Brazil selected patients with symptoms . These three studies cannot be compared with our study because their cohorts were selected among symptomatic patients or patients with achalasia-like esophagus . The remaining study by Remes-Troche et al [18] included 42 consecutive patients irrespective of their clinical symptoms . They found symptoms in 33% of the patients . The HRM assessment surprisingly showed that 28 ( 66% ) of the patients had esophageal motility disorders according to the former Chicago classification . However , with the new Chicago classification it is probable that less esophageal abnormality would have been described . Our study also found that the most frequent abnormality is minor disorders of the peristalsis , however we did not find any patients with major disorders , EGJ outflow obstruction or achalasia . On the other hand , no statistically significant relationship was seen between the presence or absence of symptoms and the esophageal HRM findings in our cohort , similarly to previous studies . [21] We acknowledge some limitations in our study . The limited sample size may have prevented us to find some significant associations between the symptoms and the manometric findings due to a possible lack of power . Moreover , a specific study with upper endoscopy or pH-metry was not performed as part of the study protocol to further characterize the heartburn referred by some of the patients . Additionally , the control group was not matched to the cohort with respect to origin and age , so some findings may be attributed to these factors . To the best of our knowledge , this is one of the few studies to date using esophageal HRM in chronic CD and the first reporting this technique in a non-endemic country . We reported esophageal motility disorder in 22 . 6% of a cohort of consecutive CD patients . Symptoms and esophagogram results did not matched with the HRM results . We believe that further prospective studies are needed , especially with HRM , to elucidate the evolution of early abnormal motility patterns in chronic esophageal CD . | Chagas disease is a parasitic disease mainly transmitted to humans by blood-sucking insects . The disease was endemic in Latin America , but it is now a global disease due to migratory movements . The disease can affect the heart and the digestive system ( mainly esophagus and colon ) . Classically , esophageal assessment in Chagas disease is performed by X-ray and self-reported symptoms . However , they lack accuracy and detect only advanced stage of the disease . Recently , new tools , such as esophageal high resolution manometry , provide more detailed information about the motility disorders of the esophagus . We assessed the esophageal involvement in patients with Chagas disease by means of high resolution manometry and compared the findings with the X-ray and self-reported symptoms . We found a low rate of mild severity motility disorders . We did not find an association between X-ray assessment and symptoms with the high resolution manometry findings . The assessment of esophageal involvement in patients with Chagas disease may benefit from early diagnosis by high resolution manometry , although more research is needed . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"biotechnology",
"medicine",
"and",
"health",
"sciences",
"classical",
"mechanics",
"laryngology",
"catheters",
"tropical",
"diseases",
"tonicity",
"parasitic",
"diseases",
"parasitic",
"protozoans",
"osmotic",
"pressure",
"hypertonic",
"dysphagia",
"organisms",
"protozoans",
"neglected",
"tropical",
"diseases",
"pressure",
"hypotonic",
"digestive",
"system",
"barium",
"chemistry",
"protozoan",
"infections",
"otorhinolaryngology",
"gastrointestinal",
"tract",
"physics",
"trypanosoma",
"cruzi",
"trypanosoma",
"chagas",
"disease",
"anatomy",
"biology",
"and",
"life",
"sciences",
"physical",
"sciences",
"chemical",
"elements",
"medical",
"devices",
"and",
"equipment",
"esophagus"
] | 2016 | High Resolution Esophageal Manometry in Patients with Chagas Disease: A Cross-Sectional Evaluation |
The ten-subunit transcription factor IIH ( TFIIH ) plays a crucial role in transcription and nucleotide excision repair ( NER ) . Inactivating mutations in the smallest 8-kDa TFB5/TTDA subunit cause the neurodevelopmental progeroid repair syndrome trichothiodystrophy A ( TTD-A ) . Previous studies have shown that TTDA is the only TFIIH subunit that appears not to be essential for NER , transcription , or viability . We studied the consequences of TTDA inactivation by generating a Ttda knock-out ( Ttda−/− ) mouse-model resembling TTD-A patients . Unexpectedly , Ttda−/− mice were embryonic lethal . However , in contrast to full disruption of all other TFIIH subunits , viability of Ttda−/− cells was not affected . Surprisingly , Ttda−/− cells were completely NER deficient , contrary to the incomplete NER deficiency of TTD-A patient-derived cells . We further showed that TTD-A patient mutations only partially inactivate TTDA function , explaining the relatively mild repair phenotype of TTD-A cells . Moreover , Ttda−/− cells were also highly sensitive to oxidizing agents . These findings reveal an essential role of TTDA for life , nucleotide excision repair , and oxidative DNA damage repair and identify Ttda−/− cells as a unique class of TFIIH mutants .
DNA-damaging agents are a constant challenge to DNA integrity . A network of DNA-repair systems collectively removes most lesions and safeguards the stability of the genome [1] . Nucleotide excision repair ( NER ) is one such DNA-repair mechanism capable of removing a wide variety of structurally unrelated DNA helix-distorting lesions , including ultraviolet light ( UV ) -induced lesions and bulky chemical adducts . Two sub-pathways have been identified: global genome NER ( GG-NER ) , eliminating distorting lesions anywhere in the genome and transcription-coupled NER ( TC-NER ) , focusing only on lesions physically blocking ongoing transcription to permit resumption of gene expression . DNA repair of helix-distorting lesions requires the helix to be opened at the site of the lesion for efficient incision of the damaged strand [2] . A protein complex essential to this process is basal transcription factor II H ( TFIIH ) . Although TFIIH was initially identified as a general RNA polymerase II transcription initiation factor [3] , this multi-subunit complex was subsequently found to have multiple functions: including RNA polymerase I transcription and , activated transcription and cell cycle control [4]–[6] . TFIIH is composed of two sub-complexes: the 7-subunit core complex comprised of xeroderma pigmentosum group B ( XPB ) , xeroderma pigmentosum group D ( XPD ) , p62 , p52 , p44 , p34 and trichothiodystrophy group A ( TTDA ) , and the associated trimeric CDK-activating kinase ( CAK ) complex involving CDK7 , MAT1 and cyclin H . Mutations in genes encoding for TFIIH subunits ( XPB , XPD and TTDA ) are associated with a surprisingly heterogeneous range of UV-sensitive clinical syndromes [7] , [8] , consistent with its diverse cellular functions . These syndromes include the ( skin ) cancer prone disorder xeroderma pigmentosum ( XP ) ; the severe neurodevelopmental and premature-aging conditions Cockayne syndrome ( CS ) and trichothiodystrophy ( TTD ) and combined forms of these syndromes , XP-CS [9] and XP-TTD [10] . TTD is a multi-systemic premature-ageing condition , characterized by brittle hair and nails , ichthyosis , and progressive mental and physical retardation [11] . Within the disease subtype known as photosensitive TTD , three TFIIH-coding genes have been found to be mutated: XPB [12] , XPD [13] , [14] and TTDA [15] . Cells isolated from TTD-A patients present a reduced amount of TFIIH , suggesting that TTDA plays an important role in stabilizing the whole TFIIH complex [15] , [16] . TTDA encodes for an 8 kDa protein that binds to the TFIIH core components XPD and p52 [17] , [18] . Although TTDA appears to be the only core TFIIH subunit that is dispensable for mammalian in vitro transcription , its presence stimulates RNA synthesis in a reconstituted transcription assay [19] . Moreover , TTDA was originally identified as a component of the yeast transcription pre-incision complex and appeared to have a role in transcription initiation in the presence of an activator [20] . TTDA resides in two cellular fractions: a TFIIH-bound fraction and a free fraction [21] . During engagement in NER , TTDA binds more tightly to TFIIH and possibly plays a role in stabilizing TFIIH on lesions , thereby facilitating the transition between subsequent NER intermediates [21] . The NER-dependent TFIIH-stabilization role can also partly restore the DNA repair deficiency seen in p52 D . melanogaster mutants ( Dmp52 ) , when an excess of TTDA molecules are available [19] . TTDA has been thought to be required only to stimulate the helix opening during NER [17] . Up to now , the function of TTDA has been presumed not to be essential for NER but only to make it more efficient [22] . Three mutations within the TTDA gene of three non-related patients with TTD have been identified [15] . Patient TTD99RO carries a homozygous transition mutation at codon 56 , converting an Arginine to a stop codon , thereby truncating the C-terminal 15 amino acids ( i . e . more than 20% ) of the protein . Patient TTD1BR is heterozygous for this allele , the other allele encodes for a transition mutation at codon 21 that converts a conserved Leucine to a Proline . Siblings TTD13PV and TTD14PV carry a homozygous mutation in the ATG start codon , aborting TTDA protein synthesis . Intriguingly , despite the rather diverse clearly severe mutations , these patients are surprisingly similar in their expression of the clinical features [15] , [23] , [24] , consistent with the idea that they all represent null-alleles . Yeast strains with a complete Ttda deletion are viable [20] , whereas complete absence of the TFIIH subunits XPB and XPD is incompatible with life in both mammals and yeast [25]–[27] , likely due to their indispensable role in transcription . Together , this suggests that TTDA is not a vital TFIIH component [15] . Primary fibroblasts from the TTD-A patients described above have been extensively examined . Although these studies have provided valuable information on the NER function of TTDA , they do not provide an accurate explanation for all the observed clinical symptoms observed in the patients . This can be explained by the fact that most disease-specific symptoms are not apparent in fibroblasts but in neuronal tissue and epithelial cells ( ichthyosis and brittle hair ) . Existing mouse models with mutations in TFIIH components strikingly mimic the clinical symptoms seen in humans [26]–[28] and have provided important information towards our understanding of the molecular and genetic bases of TFIIH-related diseases . Here we describe the generation and analysis of a Ttda knock-out ( KO ) mouse-model to investigate the molecular mechanism that leads to the TTD-specific phenotype . This investigation has shown in fact that TTDA is an essential protein for repair and embryonic development , arguing previous conclusions that needs to be modified .
To study the etiology of TTD and the cell-type-specific consequences of TTDA-deficiency , we developed a Ttda−/− mouse model . A full KO approach was considered valid , since previous evidence has indicated that TTDA is the sole core TFIIH subunit that is dispensable for viability in yeast [20] and likely also in man [15] . The targeting strategy was designed such that exon 3 ( which contains 83% of the Ttda coding sequence ) was removed and replaced by a neomycin expression cassette , driven by a PGK promoter and flanked by two LoxP sites ( referred to as LNL ) ( Figure 1A ) . Following electroporation of the linearized targeting construct into embryonic stem ( ES ) cells and G418 selection , resistant clones were screened for correct targeting by DNA blotting of NheI-digested genomic DNA with the indicated 5′probe ( Figure 1B ) . We selected two independent ES clones that had undergone homologous recombination and correct 5′ integration and presented a correct karyotype ( #J4 and #O8 ) . These clones were injected into C57bl/6 blastocysts to produce chimeric mice with germ-line transmission of the targeted Ttda allele ( referred to as Ttda+/−LNL ) . Male chimeras were mated to C57bl/6 females to produce Ttda+/−LNL heterozygous offspring . Consistent with the autosomal , recessive nature of the human syndrome the Ttda+/−LNL mice did not exhibit any obvious phenotype up to the age of 2 years . Both independent mouse lines were used for the generation of Ttda−/− mice . Since the Ttda+/−LNL mice still harbored the dominant selectable Neo marker which may interfere with the transcriptional expression of neighboring genes , the Ttda+/−LNL were crossed with the ubiquitous Cre-recombinase-expressing mouse model [29] to obtain Ttda+/− offspring . Subsequently , heterozygous mice were inter-crossed to obtain Ttda−/− mice . Genotyping of offspring was performed by PCR analysis using the sequence-specific primers shown in Figure 1A , and as indicated in Figure 1C . Surprisingly however , Ttda−/− mice were absent from the large numbers of offspring analyzed ( Table 1 and Table S1 ) . It is thus likely that homozygous loss of Ttda resulted in embryonic lethality . Additionally , the average litter size observed — when two Ttda+/− animals were intercrossed — was much lower than the litter size obtained after crossing of Ttda+/− and wild-type ( Ttda+/+ ) mice ( Table 2 ) , which again points to embryonic loss prior to full gestation . This unexpected lethality seems to contradict with the alleged non-vital function of TTDA in the human situation [15] . Moreover , this lethality is independent of the genetic background of the mouse strain used , since matings of Ttda+/− neither in C57bl/6 or in FVB background produced viable offspring . Isolated embryos from early pregnancy revealed a normal Mendelian distribution . We observed a progressive loss of phenotypically normal homozygous Ttda-deficient embryos during later stages of gestation . Ttda−/− embryos that do survive up to 19 . 5 days of gestation show delay in development and have a reduction in size and body-weight ( data not shown ) . Since we were unable to obtain Ttda−/− mice we attempted to isolate viable Ttda−/− ES cells and mouse embryo fibroblasts ( MEF ) lines . We succeeded in establishing both types of cells ruling out that the Ttda KO allele causes cellular lethality , as observed with the other TFIIH subunits XPB and XPD due to their indispensable function in basal transcription initiation . Since homozygous Ttda−/− mice were not viable , we further analyzed cells derived from Ttda−/− embryos . We determined the UV sensitivities of ES cells using a clonogenic survival assay . Surprisingly , as shown in Figure 2A , all Ttda−/− ES clones exhibit a strong UV sensitivity , to the same extent as completely NER-deficient ( Xpa−/− ) ES cells [30] . This severe UV-hypersensitivity was unexpected , since we have shown previously that TTD-A human primary fibroblasts only exhibit an intermediate UV sensitivity , caused by slow but persistent repair of UV-induced DNA lesions [22] . To find out whether this remarkable UV sensitivity in Ttda−/− ES cells is not cell type specific , we also determined the UV sensitivity of Ttda−/− mouse embryonal fibroblasts ( MEFs ) . As shown in Figure 2B , Ttda−/− MEFs were also extremely UV-sensitive to the same extent as Xpa−/− MEFs . Next , we investigated NER capacity in the Ttda−/− MEFs , by measuring: UV-induced unscheduled DNA repair synthesis ( UDS ) , i . e . a measure for GG-NER; and recovery of RNA synthesis after UV-irradiation ( RRS ) , i . e . a measure for TC-NER ( Figure 2C ) . UDS and RRS were both severely affected similar to completely NER-deficient ( Xpa−/− ) MEFs , in line with the strong UV-hypersensitivity . In addition , Ttda−/− MEFs were fully deficient in repair of 6-4 pyrimidine-pyrimidone ( 6-4PP ) photoproducts as measured by ELISA , even 16 hours after UV irradiation ( Figure 2D ) . This eliminates the option of slow but persisting repair as observed in the human TTD-A fibroblasts corroborating the complete NER-deficiency in these cells . To further dissect the stage at which this unexpected complete NER deficiency occurs in Ttda−/− MEFs , we first checked whether damage recognition is affected by Ttda ablation . As shown in Figure 2E , the damage recognizing protein Xpc is efficiently loaded to local UV-damage ( LUD ) similar as in wild-type cells . These results suggests that Ttda functions at the stage of TFIIH-loading onto XPC-bound lesions , downstream of damage recognition . To determine whether the TFIIH complex lacking Ttda was still capable of assembling at UV-induced DNA lesions — as shown in human TTD-A cells [17] , [22] — we measured the binding kinetics of TFIIH to damaged regions . In these experiments either a Multi-photon ( MP ) laser or a UV-laser was used to locally induce ( UV ) DNA damage in the nuclei of living cells [31] . To monitor TFIIH loading we used a recently developed knock-in mouse model which expresses homozygously a fluorescently-tagged ( YFP , for yellow fluorescent protein ) Xpb ( largest subunit of TFIIH ) [32] . We crossed the Ttda KO allele into this XpbYFP+f/+f background and isolated E10 . 5 MEFs from XpbYFP+f/+f Ttda+/− matings , each expressing the Xpb-YFP fusion protein . In contrast to the fast accumulation kinetics of Xpb-YFP to UV-laser damaged spots in the wild-type background ( Figure 2F and 2G ) , this protein was unable to accumulate in Ttda−/− background . Wild-type and Ttda+/− MEFs showed similar accumulation kinetics for the multi-photon ( MP ) damaged area , which induces among other lesions also UV photoproducts [31] ( Figure S1 ) . However , Xpb-YFP is incapable of accumulating to DNA damaged regions in Ttda−/− MEFs , even 15 minutes after DNA damage induction . The absence of TFIIH binding to Xpc bound DNA lesions suggests that down-stream processing of UV-lesions by the NER machinery is abrogated , since the helicase function of TFIIH is required for further assembly of the pre-incision NER complex and sequential dual incision [2] , [33] . Repair intermediates produced by NER incision induces H2AX phosphorylation ( γH2AX ) in a cell-cycle independent manner [34] , [35] . Local γH2AX after filter irradiation can be used as a sensitive marker for dual incision during NER [36] . In non-S-phase cells , since γH2AX signaling in S-phase cells is both triggered by stalled replication forks and NER . A clear local γH2AX signaling is observed in wild-type non-S-phase MEFs 1 hr after LUD , which is however absent in both Xpa−/− and Ttda−/− MEFs ( Figure 3 ) . Together these data show that Ttda is pivotal for TFIIH loading on UV lesions and that in its absence no NER-dependent dual incision occurs . We therefore consider Ttda as an essential NER factor . The absence of residual NER activity in Ttda−/− MEFs and ES cells is in contrast with the partial NER activity in TTD-A human primary fibroblasts [22] . A possible explanation for this apparent discrepancy could be derived from a human-mouse difference in NER efficiency . This could be caused by the virtual absence of CPD removal by GG-NER in rodent cells , which still occurs albeit with a slow rate in human cells [37] . Alternatively , it is not formally excluded that a mutated TTDA protein with partial biological activity is still present in the patient cells . To investigate this latter option we attempted to further reduce TTDA in human patient cells ( TTD1BR-SV ) by shRNA interference of the resident mutant TTDA transcript . The knock-down efficiency of the different shRNAs targeting the TTDA transcript was verified by RT-qPCR ( Figure 4A ) . We selected two shRNAs ( #3398 and #3402 ) that were most efficient in reducing the resident TTDA transcript to approximately 5% of the initial amount . Next , we determined the UV-sensitivity by performing a clonogenic survival assay ( Figure 4B ) . Surprisingly , depleting mutant TTDA mRNA in TTD-A human patient cells severely aggravated the UV-sensitivity , whereas a control non-targeting shRNA did not have any effect . These results suggest that the mutated human TTDA proteins still harbor residual NER activity and that the severe NER-deficient phenotype observed in Ttda−/− cells is not specific for mouse cells . To find out whether this strong NER deficiency upon TTDA depletion by shRNA in the human cells was due to a further reduction in cellular TFIIH content below a critical threshold , we determined TFIIH levels in the parental TTD-A cells and their cognate shRNA-TTDA-depleted cells . As shown by the immunofluorescence staining of the XPB subunit of TFIIH in Figure 4C and Figure S2 there was no further decline of the already low TFIIH levels upon TTDA depletion by shRNA in TTD-A cells . Since depletion of mutant TTDA aggravated UV-sensitivity , we further investigated the functionality of the mutant human proteins ( schematically depicted in Figure 5A ) . Because TTDA seems to be required for efficient loading of additional NER factors , the ability of TTDA to localize to LUD is indicative of its function in NER . To determine the binding of mutant TTDA to laser-induced LUD in living cells , we transduced Ttda−/− MEFs with lentiviruses encoding for GFP-tagged TTDA: TTDAWT-GFP ( wild type ) , TTDAM1T-GFP ( start site mutation , using the first downstream Methionine codon ( M16 ) ) , TTDAR56X-GFP ( premature stop mutation ) or TTDAL21P-GFP ( transition mutation ) . The latter three mimic mutations found in TTD-A patients . TTDA-GFP fully complements the UV-sensitivity of Ttda−/− MEFs ( Figure S3A and [21] ) and shows that the tagged TTDA is biologically active . As shown in Figure 5B and 5C , TTDAR56X-GFP and TTDAL21P-GFP accumulated with similar initial kinetics as TTDAWT-GFP to LUD , though prior to steady-state less TTDAL21P-GFP accumulated compared to TTDAWT-GFP . This would suggest that the binding time of this mutant protein is reduced compared to the wild-type protein . The ability to accumulate at DNA damaged regions was also confirmed by an immunofluorescence experiment , using a 5 µm filter and a UV-C lamp to apply local UV damage ( Figure S3B ) . However , we did not find accumulation of TTDAM1T-GFP , despite the fact that patient cells carrying this mutation display only a mild NER-deficient phenotype [15] . This apparent discrepancy could be explained by a possible combinational functional interference of both the 16 amino acid N-terminal truncation and the C-terminal GFP-tag , despite the notion that over expression of a full length TTDA-GFP rescues the UV-sensitivity of Ttda−/− MEFs ( Figure S3A ) . To further investigate a possible partial function of the translational start-site mutant , we over expressed non-tagged mutant TTDA cDNA — mimicking the translation start mutation found in TTD13 PV and TTD14 PV patients — and the other mutants in Ttda−/− MEFs and assayed for UDS . Co-transfected GFP served as a marker to identify transfected cells . Over expression of all TTDA mutant cDNAs clearly corrected the DNA repair synthesis deficiency of Ttda−/− MEFs to almost wild-type levels , including the translational start-site ( M1T ) mutant ( Figure 5D , 5E ) . Together these results strongly suggest that TTD-A patients do express a partially functional mutant protein and confirms that also in human cells TTDA is an essential NER component , rather than only an NER accessory factor . The new finding of TTDA's essential role in NER has far-reaching biological significance . However , the complete NER deficiency cannot explain the embryonic lethality , since other mice with fully compromised NER function ( such as Xpa−/− mice ) do not display similar developmental abnormalities and are viable [38] , [39] . To gain further insight into the nature of this lethal phenotype , we analyzed whether the altered genomic locus of the Ttda KO allele interferes with the expression of neighboring genes . The deletion of genomic sequences in the Ttda gene may include cryptic or unrecognized transcriptional enhancers or insulators which may create hypomorphic expression of adjacent genes ( see [40] for a typical example ) . To investigate possible transcriptional interference , we analyzed the expression levels of the 3 most proximal neighboring genes to the Ttda gene in Ttda−/− ES cells by quantitative real-time PCR ( RT-qPCR ) : Synaptojanin 2 ( Synj2 , 35 . 5 Kb upstream ) , Serine active site containing 1 ( Serac1 , 89 bp upstream ) and Tubby-like protein 4 ( Tulp4 , 21 . 3 Kb downstream ) . The expression of these genes was not reduced when compared with expression in their heterozygous and wild-type cognates ( Figure 6A ) . Two of these genes showed increased expression in the Ttda−/− cells . This likely does not cause embryonic lethality , since cells from heterozygous animals , which also had this increased expression do not display any obvious phenotype . Expression analysis of the Ttda allele neighboring genes at a later stage of development ( E11 . 5 ) ( Figure S4 ) confirmed the absence of a clear correlation between aberrant gene expression and lethality in Ttda−/− embryos . It has been suggested that the TTDA protein is a repair-specific TFIIH-subunit , which is not strictly required for basal transcription as the other TFIIH components are [17] . However , mutated TTDA causes an overall reduction of TFIIH protein abundance in human fibroblasts [16] . This sub-limiting amount of TFIIH does not cause a significant reduction of basal transcription within cultured TTD-A patient fibroblasts [16] . On the other hand , in the developing mouse embryo a decreased amount of TFIIH might reduce the transcription capacity which is needed to produce a fully developed animal . To directly monitor the quantity of TFIIH in living cells , we isolated ES cells and E10 . 5 MEFs from XpbYFP+f/+f Ttda+/− matings ( see above ) , each expressing the XPB-YFP fusion protein . A clear reduction in the quantity of TFIIH ( as deduced from the strong reduction of the YFP signal ) was easily observed in the live cell images of both Ttda KO ES cells and MEFs ( Figure 6B–6C ) . We determined by direct fluorescence measurements the YFP signal emitted from the nuclei of ES cells and MEFs , which was respectively 22% and 33% in the KO cells as compared to wild-type cells isolated from litter mates . These levels are for the MEFs comparable to the amounts measured in human cells ( approximately 30% ) [16] and appeared even lower in ES cells . Next we analyzed the transcription capacity in these murine cells by pulse labeling de novo RNA synthesis for either 30 minutes or 2 hours with fluorescent based 5-ethynyl-uridine ( EU ) . The average fluorescence intensity , which is a measure for the total amount of transcription in these cells , was evaluated ( Figure 6D ) . In accordance with the reduced steady-state level of TFIIH , measured in ES cells and MEFs , the overall transcription appeared to be significantly reduced in Ttda−/− cells as well . Despite the severely reduced amount of TFIIH in Ttda−/− cells , which also attenuated overall transcription , the proliferative capacity of the embryonic cells was not affected ( Figure S5A ) . It is thus not likely that reduced overall transcription would be the sole cause to the observed embryonic lethality . Detailed analysis of different mutant TFIIH mouse models revealed a correlation between sensitivity to oxidative DNA damage and severity of the phenotype of the different models . These observations argued for a unknown function of TFIIH in oxidative lesion removal [27] . To investigate whether Ttda−/− cells are also defective in repairing other ( non NER-type ) DNA lesions , we measured their sensitivity to several oxidizing agents . Clonogenic survival assays performed on Ttda−/− ES cells revealed hyper-sensitivity to gamma irradiation ( Figure 7A ) and potassium bromate ( Figure 7B ) , similar to Csb−/− ES cells ( known to be sensitive to oxidative DNA damage ) [30] . Since the Xpa−/− ES cells assayed in parallel were not sensitive to any of these agents , this sensitivity is not a general effect of NER-deficiency . This phenomenon is also not cell-type specific , since Ttda−/− MEFs are also hyper-sensitive to gamma irradiation ( Figure 7C ) . To exclude the possibility that Ttda−/− cells have a general low tolerance to DNA damage , we also measured their sensitivity to mitomycin C ( MMC ) . MMC induces inter-strand cross-linking that is specifically repaired by the inter-strand cross-link repair pathway — a pathway which specifically involves the NER protein complex XPF-ERCC1 ( other NER factors are not required to remove this class of lesions ) [41] . As shown in Figure 7D , Ercc1−/− cells are highly sensitive to MMC treatment . All the other NER mutant cells , including the Ttda−/− cells , are not sensitive . Since most oxidative DNA lesions are removed by the base excision repair ( BER ) genes [42] , we wondered whether the oxidative DNA damage hypersensitivity could be due to reduced expression of BER genes , as a consequence of the low TFIIH level in Ttda−/− cells . To that aim we analyzed the expression of the core BER genes by RT-qPCR , since the absence of single oxidative damage-specific glycosylases does not cause cellular hyper-sensitivity due to ( partial ) redundant glycosylases [43] . As shown in Figure S5B , none of the BER genes were lower expressed in the Ttda−/− cells . To further investigate a possible general BER defect in Ttda−/− cells , we tested for alkylating DNA damage sensitivity . Apart from the initial recognizing glycosylases , further processing of these lesions follows the same route as for oxidative DNA damage . To that aim we treated the cells with varying concentrations of Methyl methanesulfonate ( MMS ) . In contrast to oxidative DNA damage , Ttda−/− appeared not hypersensitive to this agent ( Figure 7E ) . Together our data unambiguously establish a function for TTDA and likely for the entire TFIIH complex in the tolerance to oxidative DNA damage .
Analysis of NER parameters in embryonic Ttda−/− cells revealed a remarkably severe NER-deficiency and includes extreme UV-hypersensitivity , absence of UV-induced UDS and defective in removing UV-induced lesion ( comparable to Xpa−/− cells ) . This complete NER-deficient phenotype is in striking contrast to the only mild NER defect seen in human TTD-A cells . The NER-deficiency observed cannot be explained by a cell-type specific UV-response , since MEFs and ES cells are equally UV-sensitive . In a previous study we noted very low levels of all subunits of TFIIH in human TTD-A patient cells and suggested that this contributes to the associated partial NER defect [16] . MEFs derived from a previously generated TTD mouse model [26] — mimicking a known human TTD-causative point mutation ( R722W ) in the mouse Xpd locus — are only slightly sensitive to UV irradiation and exhibit a mild UDS defect and have decreased TFIIH levels [28] . Therefore , complete NER-deficiency appears to be specific to the Ttda−/− cells and not a general TTD-associated phenotype caused by a lower level of TFIIH . Since the NER-phenotype of Xpd TTD mice closely mimics the partial repair deficiency features in human XP-D TTD cells , it is thus unlikely that this discrepancy in the severity of the NER-phenotype between man and mice is a species-specific phenomenon of increased NER-deficiency in murine cells . Our dynamic in vivo studies have revealed that the TTDA defect is located at a stage prior to stable association of TFIIH with the NER initiation complex , containing XPC ( Figure 2E ) . However , in contrast to NER , TTDA seems dispensable for loading TFIIH onto promoter sequences for transcription initiation of RNA pol I and II , in view of the fact that Ttda−/− cells are viable and do not show a proliferation defect . Knock-down of the resident TTDA transcripts in TTD-A patient cells rendered these cells extremely UV-sensitive , with the same level of sensitivity as completely NER-deficient XP-A cells and similar to the Ttda−/− MEFs . From these data we conclude that in humans the mutant TTDA protein are partially functional . This hypothesis is further substantiated by the finding that the GFP-tagged TTDAR56X protein ( homozygous mutation in patient TTD99RO ) was able to accumulate at sites of UV-induced damages , in line with the suggested presence of residual activity of mutant TTDA proteins in patient TTD-A cells . Our data suggest that the TTDA protein in TTD13/14PV cells is partially functional . This observation is particularly intriguing , since it has been suggested that no TTDA protein is expressed in these cells , due to the homozygous translational start-site mutation [15] . However , the fact that depletion of the mutant TTDA mRNA by targeted shRNA interference further enhanced UV sensitivity , suggests that this mutant mRNA is still able to generate a partially functional TTDA protein . One way to explain this phenomenon is by assuming that despite this ATG mutation , some TTDA protein is still being produced by initiating from an in-frame downstream ATG ( codon 16 ) . The usage of such alternative start site may produce low levels of an N-terminally truncated TTDA protein , sufficient to rescue lethality and complete NER-deficiency . Indeed , over expression of a 16 amino-acid N-terminally truncated mutant TTDA , rescued the UDS defect of Ttda−/− cells ( Figure 5D , 5E ) and argues for partial functional production of N-terminally truncated TTDA in TTD12/14PV patient cells . Interestingly , it has been shown ( both in yeast and humans ) that the N-terminal domain of TTDA is important for binding to the TFIIH subunits XPD and p52 [17] , [44] and for stimulating the ATPase-activity of the XPB subunit [17] . Furthermore , this interaction is also critical for the TFIIH stability . Apparently , the truncated Ttda protein is nevertheless able to carry out part of its function to permit residual NER . Defects in multiple DNA repair systems may cause synergistic effects or even synthetic lethality [45] . For example , severe developmental and premature aging problems have been seen in KO mouse models of DNA repair factors that function in independent repair pathways , such as ERCC1 ( which functions in NER and inter-strand cross-link repair ) and Xpa−/− Csbm/m double KOs ( which is defective in GG-NER , and presumably also in the broad TCR pathway ) [46] , [47] . It has been suggested that endogenously produced DNA lesions ( e . g . from reactive oxygen species ( ROS ) or lipid peroxidation byproducts ) that cannot be removed because of the repair defect , are in part responsible for the phenotype observed . Here we have shown that Ttda−/− cells are sensitive to several oxidizing agents to the same extent as Csb−/− ES cells [30] . Based on these results , we suggest that Ttda−/− embryos are confronted with unrepaired endogenous oxidative lesions , possibly generated by low but continuous exposure to ROS during development . This compromised repair of oxidative DNA lesions may contribute to the observed lethal phenotype observed in the Ttda−/− embryos . Importantly , this reduced resistance to oxidative DNA damage is likely not caused by a general ( core ) BER defect . Previously , it was suggested that TFIIH is implicated in coordinating incision of lesion-stalled transcription complexes [48] and that some oxidative DNA lesions are processed by transcription-coupled repair [49] . It is thus possible that TTDA ( TFIIH ) is involved in a specific — thus far uncharacterized — transcription-coupled repair process of oxidative DNA damage . The fact that TTDA has a pivotal role not only in NER but also in oxidative DNA damage defense and transcription , argues for a function of TTDA in various DNA metabolizing processes , causing synergistic effects when inactivated . However , full NER-deficiency in combination with defects in oxidative DNA damage repair — as in Xpa−/− Csb−/− double KO mice — does not lead to embryonic lethality . In this mouse model pups are born , but they progressively develop very severe neurologic symptoms and premature aging features . It is thus likely that not only the repair functions contribute to the lethal phenotype of Ttda−/− animals , but that also its function in transcription is involved . Nevertheless , human TTD-A patient derived cells also express low levels of TFIIH , but have only limited post-natal developmental problems [23] . It is possible that transcription is more demanding during mouse embryogenesis than in the human embryo , due to its more rapid development . In this respect it should be noted that TTD-causing mutations in the human XPD gene are associated with impaired placental development and other gestational complications [50] . It has been shown that TTDA is dispensable for mammalian in vitro transcription [17] . Nevertheless , TTDA was originally found to be present in the pre-initiation complex [20] and reconstituted transcription assays have demonstrated that TTDA stimulates this reaction [19] . It has been hypothesized that mutations affecting XPD's function in DNA repair cause the disorder XP — associated with a 1000-fold increased risk of skin cancer — while mutations affecting XPD's role in RNA Polymerase II-mediated transcription lead to TTD-specific features: brittle hair and nails , and scaly skin [51] . TTD features in TTD-A patients are relatively mild compared to those seen in XPD-associated TTD patients . The mild TTD-phenotype suggests that the role of TTDA in transcription is plausible , but also that it is not the only cause for embryonic lethality . We have demonstrated that Ttda KO cells have a low steady state level of TFIIH and accordingly have a lower transcriptional activity ( Figure 6C and 6D ) . The low TFIIH quantity does not seem to be the sole cause of embryonic lethality , since similar low levels of TFIIH are observed in TTD-A patient cell lines , which are compatible with life . However , the notion of reduced transcriptional activity in Ttda−/− cells argues that this feature may contribute to embryonic lethality . For instance , during certain stages of embryonic development which requires high transcriptional capacity , normal embryogenesis may be compromised . Moreover , it cannot be excluded that mutations in TTDA affect the transcription of a subset of specific genes , as shown in cells with XPD-associated TTD mutations defective in activated-transcription of nuclear receptors [52] . In this scenario , the expression of specific genes , essential for development of the embryo might be disturbed hindering proper embryogenesis and finally inducing in utero death . In both cases , TTDA function appears to extend beyond the previously suggested main function in NER , as it is also important for both development and viability . Our data clearly show that TTDA has an essential function in NER . The rather mild TTD-phenotype observed in TTD-A patients is due to the presence of partly functional mutant proteins . The sensitivity to endogenously produced oxidative DNA lesions in Ttda−/− cells suggests that TTDA ( and likely the entire TFIIH ) has additional functions in DNA repair extending beyond NER . The lethal phenotype observed in Ttda−/− embryos is possibly the result of several defects , such as insufficient levels of TFIIH needed for transcription in highly proliferative tissues , impairment in the activated transcription of specific genes , and unrepaired lesions — induced either by UV or endogenously by oxidizing agents which are relevant for cancer as well as aging .
All animal work was conducted according to the Federation of European Laboratory Animal Science Associations ( FELASA ) ethical requirements and with respect of the 3R animal welfare rules . The knock-out targeting vector ( backbone Puc18 ) contained a 12 . 5 Kb XbaI fragment of mouse genomic DNA ( isogenic to 129Ola ) harboring the entire Ttda locus . The complete exon 3 ( i . e . most of the protein coding sequence ) was excised , using BalI digestion , and replaced with a neomycin gene-expression cassette , flanked by two LoxP sites [53] and used as a dominant selectable marker . The dominant marker was inserted in the same transcriptional orientation as the Ttda gene . ES cells ( 129Ola , subclone IB10 ) were cultured in BRL-conditioned medium supplemented with 1 , 000 U/ml leukemia inhibitory factor ( LIF ) . A total of 20 µg of the SalI linearized targeting vector was electroporated into approximately 107 ES cells in 500 µl . Selection with 0 . 2 µg/ml G418 was started 24 h after electroporation . G418 resistant clones were isolated after 8–10 days . Screening for homologous recombinants was performed using DNA blot analyses of NheI digested DNA with a 1000 bp 5′ external probe A ( see Figure 1A ) . 13 ES clones out of 130 G418-resistant clones , had a correctly targeted Ttda allele . Two of the 13 correctly targeted ES clones were checked for proper karyotype and injected into blastocysts from C57bl/6 mice and then transplanted into B10/CBA foster mothers . Chimeric mice were further crossed , and germ line transmission of the targeted allele to offspring was genotyped by PCR ( Figure 1 ) . Primer sequences are available on request . Total RNA was purified from ES cells and 11 . 5-day-old embryos using the RNeasy Mini Kit ( Qiagen ) . cDNA was created from 2 µg of RNA using an RT kit ( Invitrogen ) and random primers ( Invitrogen ) . 5 µl of this cDNA was used in the following reaction with 29 nM sense and antisense primer ( primer sequences available on request ) . The cDNA was amplified by real-time quantitative PCR ( RT-qPCR ) in 25 µl reactions using platinum Taq polymerase and SYBR green according to manufacturer's protocol ( Invitrogen ) with the c1000 Thermal Cycler ( Bio-Rad ) . Reaction conditions were: 95°C for 3 min , followed by 40 cycles of 95°C for 15 sec , 60°C for 30 s and 72°C for 30 sec and ending with 95°C for 1 min . Expression levels were normalized to Gapdh using the method described by Pfaffl [54] . Specificity of the reactions was confirmed by analysis of the RT-qPCR melt curves . ES cells and MEFs were grown in a 6 cm dish and cultured for 2 days prior to the experiments . The cells were washed once with PBS and incubated for 30 minutes or 2 hours in culture medium containing 0 . 1 µM 5-ethynyl-uridine ( EU ) . After EU incorporation , cells were treated in the same way as described above ( UDS; fluorescent assay ) . 70–80% confluent cultures of MEFs to be analyzed were washed with PBS , UV irradiated ( 8 J/m2 ) and incubated for various time points ( 1 to 16 hours ) . Cells were harvested in PBS and DNA was isolated using QIAamp DNA Blood Mini Kit ( QIAGEN ) . DNA concentrations were determined by measuring the optical density at 260 nm . 96-well polyvinylchloride flat-bottom micro titer plates were precoated with 0 . 003% Protamine Sulfate , 50 µl/well and dried in the dark overnight at 37°C ( Sigma ) . DNA samples were denatured for 10 minutes at 95°C and immediately cooled on ice for 20 min . 50 µl/well of vortexed DNA solution in H2O was loaded in the precoated 96-well plate to a final concentration of 6 µg/ml for the detection of 6-4PP . The plate was dried overnight at 37°C , then washed 5 times with PBS+0 . 05% Tween-20 ( 150 µl/well ) . The wells were pre-absorbed with PBS+2% FCS for 30 minutes at 37°C and subsequently washed 5 times with PBS/Tween-20 prior to incubation for 30 min at 37°C with 100 µl/well of primary antibody: 6-4PP ( 1∶1000; Bioconnect/MBL ) diluted in PBS . After 5 washes with PBS/Tween-20 , samples were incubated for 30 minutes at 37°C with 100 µl/well of secondary antibody: goat anti-mouse IgG ( H+L ) conjugated to HRP ( 1∶2000; Southern Biotech ) . After 5 washes with PBS/Tween-20 , samples were treated with 100 µl/well Citrate-phosphate buffer ( 24 mM C6H8O7 . H2O and 41 mM Na2HPO4 . 2H2O; Sigma ) . Samples were then incubated with 100 µl/well of freshly made ODP buffer ( 0 . 4% o-Phenylene damine and 0 . 02% citrate-phosphate buffer; Sigma ) at 37°C for 30 min . The reaction was stopped by adding 50 µl/well of 2 M H2SO4 and absorbance was immediately measured at 490 nm . Cells were grown on glass cover slips ( 24 mm ) for three days prior to the experiments and fixed with 2% paraformaldehyde ( Sigma ) at 37°C for 15 minutes . Cover slips were washed three times for 5 minutes with PBS containing 0 . 1% Triton X-100 ( Sigma ) . To visualize the DNA photoproducts , nuclear DNA was denatured by incubation with 0 . 07 N NaOH ( Sigma ) at room temperature for 5 minutes and washed tree times for 5 minutes with 0 . 1% Triton X-100 and subsequently for 5 minutes with PBS+ ( PBS containing 0 . 15% glycine ( Sigma ) and 0 . 5% BSA ( Sigma ) ) . Cells were incubated at room temperature with primary antibodies for 2 hrs in a moist chamber . Subsequently , cover slips were washed three times with PBS/Triton X-100 and PBS+ , incubated 1 hour with secondary antibodies at room temperature and again washed three times in PBS/Triton X-100 . Samples were embedded in Vectashield mounting medium ( Vector Laboratories ) . Images were obtained using a confocal laser scanning microscope ( LSM 510; Zeiss ) . The primary antibodies used for this procedure were mouse anti-CPD ( 1∶1000; TDM-2; BioConnect/MBL ) , rabbit anti-GFP ( 1∶1000; ab290; Abcam ) , rabbit anti XPB ( 1∶1000; S19; Santa Cruz ) , rabbit anti XPC ( 1∶200; fraction 5; home-made ) , mouse anti γH2AX ( 1∶1000; 07-164; Millipore ) and mouse anti-MDC1 ( 1∶1000; MDC1-50; Abcam ) . The secondary antibodies used were Alexa Fluor 594 goat anti-mouse IgG ( H+L ) ( 1∶1000; Molecular probes ) , Alexa Fluor 594 goat anti-rabbit IgG ( H+L ) ( 1∶1000; Molecular probes ) , Alexa Fluor 488 goat anti-rabbit IgG ( H+L ) ( 1∶1000; Molecular probes ) , Alexa Fluor 488 goat anti-mouse IgG ( H+L ) ( 1∶1000; Molecular probes ) and Alexa Fluor 647 azide ( 1∶400; invitrogen ) . To measure proliferation in MEFs , equal number of cells from early passages ( p2 ) were plated in 6 cm culture dishes ( approximately 1×104 cells per well ) in triplicate in 3 ml medium ( day 0 ) . The medium was changed every 2 days cells and cells were counted 1 , 3 , 5 and 6 days after seeding , using a cell counter ( Beckman Coulter Z2 ) . | DNA is under constant attack of various environmental and cellular produced DNA damaging agents . DNA damage hampers normal cell function; however , different DNA repair mechanisms protect our genetic information . Nucleotide Excision Repair is one of the most versatile repair processes , as it removes a large variety of DNA helix-distorting lesions induced by UV light and various chemicals . To remove these lesions , the DNA helix needs to be opened by the transcription/repair factor II H ( TFIIH ) . TFIIH is a multifunctional complex that consists of 10 subunits and plays a fundamental role in opening the DNA helix in both NER and transcription . TTDA , the smallest subunit of TFIIH , was thought to be dispensable for both NER and transcription . However , in this paper , we show for the first time that TTDA is in fact a crucial component of TFIIH for NER . We demonstrate that Ttda−/− mice are embryonic lethal . We also show that Ttda−/− mouse cells are the first known viable TFIIH subunit knock-out cells , which are completely NER deficient and sensitive to oxidative agents ( showing a new role for TFIIH outside NER and transcription ) . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"animal",
"models",
"rna",
"model",
"organisms",
"nucleic",
"acids",
"mouse",
"genetics",
"molecular",
"genetics",
"dna",
"biology",
"dna",
"repair",
"molecular",
"cell",
"biology",
"rna",
"synthesis",
"genetics",
"of",
"disease",
"gene",
"function"
] | 2013 | Disruption of TTDA Results in Complete Nucleotide Excision Repair Deficiency and Embryonic Lethality |
We previously reported that the G allele of rs3853839 at 3′untranslated region ( UTR ) of Toll-like receptor 7 ( TLR7 ) was associated with elevated transcript expression and increased risk for systemic lupus erythematosus ( SLE ) in 9 , 274 Eastern Asians [P = 6 . 5×10−10 , odds ratio ( OR ) ( 95%CI ) = 1 . 27 ( 1 . 17–1 . 36 ) ] . Here , we conducted trans-ancestral fine-mapping in 13 , 339 subjects including European Americans , African Americans , and Amerindian/Hispanics and confirmed rs3853839 as the only variant within the TLR7-TLR8 region exhibiting consistent and independent association with SLE ( Pmeta = 7 . 5×10−11 , OR = 1 . 24 [1 . 18–1 . 34] ) . The risk G allele was associated with significantly increased levels of TLR7 mRNA and protein in peripheral blood mononuclear cells ( PBMCs ) and elevated luciferase activity of reporter gene in transfected cells . TLR7 3′UTR sequence bearing the non-risk C allele of rs3853839 matches a predicted binding site of microRNA-3148 ( miR-3148 ) , suggesting that this microRNA may regulate TLR7 expression . Indeed , miR-3148 levels were inversely correlated with TLR7 transcript levels in PBMCs from SLE patients and controls ( R2 = 0 . 255 , P = 0 . 001 ) . Overexpression of miR-3148 in HEK-293 cells led to significant dose-dependent decrease in luciferase activity for construct driven by TLR7 3′UTR segment bearing the C allele ( P = 0 . 0003 ) . Compared with the G-allele construct , the C-allele construct showed greater than two-fold reduction of luciferase activity in the presence of miR-3148 . Reduced modulation by miR-3148 conferred slower degradation of the risk G-allele containing TLR7 transcripts , resulting in elevated levels of gene products . These data establish rs3853839 of TLR7 as a shared risk variant of SLE in 22 , 613 subjects of Asian , EA , AA , and Amerindian/Hispanic ancestries ( Pmeta = 2 . 0×10−19 , OR = 1 . 25 [1 . 20–1 . 32] ) , which confers allelic effect on transcript turnover via differential binding to the epigenetic factor miR-3148 .
Systemic lupus erythematosus ( SLE [OMIM 152700] ) is a complex and heterogeneous autoimmune disease with a strong genetic component that is modified by environmental exposures . Although the detailed etiopathogenesis of SLE remains unknown , excessive innate immune activation involving toll-like receptors ( TLRs , particularly TLR7/8/9 ) and type I interferon ( IFN ) has been recognized as an important pathogenic mechanism in the disease [1] . Therapeutics targeting the TLR/IFN pathway are in development for the treatment of SLE , with ongoing clinical trials investigating monoclonal antibodies against IFN-α and inhibitors for TLR7/TLR9 ( reviewed in [2] ) . Recent genome-wide association ( GWA ) and follow-up studies have revealed the association of a number of polymorphic variants in genes encoding components of the TLR/type I IFN pathway with susceptibility to SLE ( reviewed in [3] , [4] ) , providing insights at the molecular level to refine our understanding of this dysregulated pathway in the predisposition to SLE . Our previous study identified a single nucleotide polymorphism ( SNP ) , rs3853839 , in the 3′ UTR of an X-linked gene TLR7 to be associated with SLE in 4 , 334 cases and 4 , 940 controls of Eastern Asian descent [5] , providing the first convincing evidence for the genetic contribution of TLR7 to human SLE . Individuals carrying the risk G allele exhibited increased TLR7 transcripts and a more robust IFN signature than non-risk C allele carriers [5] . In this study , by fine mapping the TLR7-TLR8 region , we confirmed that the previously reported functional SNP rs3853839 , located within a predicted binding site of miR-3148 , was most likely responsible for observed association with SLE in three populations of non-Asian ancestry . We demonstrated a differential miR-3148 modulation explaining the effect of allelic variation at rs3853839 on TLR7 expression .
We conducted genotyping and imputation for genetic variants covering ∼80 kb of the TLR7-TLR8 region on Xp22 . 2 . After applying quality control measures , 41 genotyped SNPs and 57–75 imputed SNPs/INDELs ( insertion-deletion ) ( varying among different ancestries ) were assessed for association with SLE in unrelated cases and healthy controls of European American ( EA , 3 , 936 cases vs . 3 , 491 controls ) , African American ( AA , 1 , 679 vs . 1 , 934 ) and Hispanic enriched for the Amerindian-European admixture ( HS , 1 , 492 vs . 807 ) descent ( Figure 1A ) . The strongest association signal was consistently detected at rs3853839 in the three ancestries , including EA ( minor allele frequency of 20 . 3% in cases vs . 17 . 2% in controls , P = 6 . 5×10−6 , OR [95%CI] = 1 . 23 [1 . 13–1 . 35] ) , AA ( 19 . 8% vs . 16 . 7% , P = 1 . 1×10−3 , OR = 1 . 24 [1 . 09–1 . 41] ) and HS ( 44 . 8% vs . 37 . 3% , P = 7 . 5×10−4 , OR = 1 . 26 [1 . 10–1 . 43] ) ( Figure 1B , Table 1 ) . After Bonferroni correction for multiple comparisons , the association of rs3853839 with SLE remained significant in EA and HS , and reached a nominal significance in AA . Combining the EA , AA and HS datasets , the meta-analysis P value of rs3853839 ( Pmeta = 7 . 5×10−11 , OR = 1 . 24 [1 . 18–1 . 34] ) exceeded the commonly used threshold of 5×10−8 for genome-wide significance ( Figure 1C , Table 1 ) . Thus , the association of rs3853839 with SLE previously identified in Eastern Asians was confirmed in three non-Asian ancestries . Only six other SNPs within a relatively small interval of 5 kb spanning from TLR7 3′downstream to TLR8 intron 1 were consistently associated with SLE ( P<0 . 05 ) in EA , AA and HS ( Table S1 ) , and remained significant trans-ancestral meta-analysis P values after Bonferroni correction ( 5 . 5×10−6≤Pmeta≤1 . 3×10−6 , Table S1 ) . Linkage disequilibrium ( LD ) analysis revealed low LD strength between rs3853839 and these SNPs across non-Asian ancestries ( r2<0 . 26 , 0 . 37 , and 0 . 51 in EA , AA and HS , respectively ) , but these 6 SNPs are in strong LD with each other and could be defined as a block ( Figure S1 ) . Among them , non-synonymous SNP rs3764880 ( Met1Val ) located at TLR8 exon1 exhibited the strongest association ( Pmeta = 1 . 3×10−6 , OR = 1 . 15; Table S1 ) . To distinguish whether the associations of these 6 SNPs with SLE were independent of rs3853839 , we performed conditional haplotype-based association test . After conditioning on rs3853839 , association signals detected at these 6 loci were completely eliminated in EA , AA and HS ( Figure S1 ) . In contrast , conditioning on rs3764880 , a consistent association signal was detected at rs3853839 in EA and HS ( Figure S1 ) , indicating that the association signals detected at these 6 SNPs might be attributed to that of rs3853839 . Taken together , we confirmed rs3853839 as the only SNP in the TLR7-TLR8 region showing an independent association with SLE across all three non-Asian ancestries . A meta-analysis by combining all datasets of Asian and non-Asian ancestries showed compelling evidence of association with SLE at rs3853839 ( Pmeta = 2 . 0×10−19 , OR = 1 . 25 [1 . 20–1 . 32] , Table 1 ) . Given the location of TLR7 at X chromosome , we examined the allelic association of rs3853839 separately by gender . Of note , the sex-specific association of rs3853839 with SLE previously detected in Asian men [5] was not replicated in non-Asian ancestries ( Table 1 ) . Given the convincing evidence for the trans-ancestral association of rs3853839 with SLE susceptibility , we then evaluated its effect on regulation of TLR7/8 expression . Messenger RNA ( mRNA ) levels of TLR7 and the two alternative TLR8 isoforms were measured by real-time PCR in PBMCs from healthy EA individuals ( n = 62 ) . TLR7 mRNA levels were significantly different among women ( n = 41 ) carrying different genotypes of rs3853839 [P = 0 . 003 , one-way analysis of variance ( ANOVA ) ] , in which the GG and GC carriers exhibited notably increased TLR7 mRNA levels compared with the CC carriers [P = 0 . 02 for GG ( n = 5 ) vs . CC ( n = 18 ) and 0 . 02 for GC ( n = 18 ) vs . CC , respectively , Student's t test; Figure 2A ) and the number of rs3853839 risk G allele was significantly correlated with increased TLR7 mRNA levels ( R2 = 0 . 26 , P = 8×10−4 , linear regression test ) . Consistently , male G allele carriers ( n = 5 ) also had significantly higher TLR7 mRNA expression than male C allele carriers ( n = 16 ) ( P = 0 . 01 , Figure 2A ) . There was no significant association of rs3853839 genotypes with mRNA levels of two TLR8 isoforms in either women or men ( Figure 2A ) . No sex differences in TLR7 or TLR8 mRNA levels were observed between individuals carrying the same genotype [GG women vs . G men: P = 0 . 41 ( TLR7 ) , 0 . 63 ( TLR8a ) and 0 . 50 ( TLR8b ) ; CC women vs . C men: P = 0 . 10 ( TLR7 ) , 0 . 91 ( TLR8a ) and 0 . 65 ( TLR8b ) ] . These results were in accordance with our previous observations in Chinese [5] , supporting the importance of rs3853839 in regulating TLR7 rather than TLR8 gene expression . We assessed the intracellular expression of TLR7 and TLR8 proteins by flow cytometry in PBMCs from 7 pairs of healthy women ( GG vs . CC ) and men ( G vs . C ) , respectively . Of the 7 pairs of individuals in each gender , 4 pairs were of EA descent and 3 pairs were Asians . Compared with C allele carriers , G allele carriers had significantly higher TLR7 protein levels in PBMCs ( P = 0 . 038 and 0 . 009 in women and men , respectively; Figure 2B ) , especially in CD19+ B cells and CD14+ monocytes ( Figure S2 ) . No significant association between rs3853839 genotypes and TLR8 protein levels was observed in either total PBMCs or in specific cell subsets ( Figure S3 ) . We next performed luciferase reporter assays to further confirm the functional effect of rs3853839 on TLR7 expression . PCR-amplified TLR7 3′UTR fragments with either the G or C allele of rs3853839 were cloned downstream of an SV40 promoter-driven Renilla luciferase gene in the psiCHECK-2 vector , which also contained a firefly luciferase gene to serve as an internal transfection normalization control ( Figure 2C ) . Constructs were then transiently transfected into either HEK-293 or differentiated HL-60 ( dHL-60 , neutrophil-like cells ) cells . After 24 hours , cell lysates transfected with the G-allele construct showed significantly higher luciferase activity than those transfected with the C-allele construct in both HEK-293 and dHL-60 cells ( P = 0 . 026 and 0 . 009 , respectively; Figure 2C ) . Taken together , consistent results from ex vivo and in vitro studies indicated that the SLE-risk G allele of rs3853839 conferred elevated TLR7 expression at the both mRNA and protein level . To explore the mechanism of rs3853839 in regulating TLR7 mRNA turnover , we assessed allelic difference in TLR7 mRNA degradation by pyrosequencing . We first determined the rs3853839 G/C allele ratio in genomic DNA ( gDNA ) and cDNA from healthy EA women ( n = 7 ) carrying the GC genotype . The mean G/C allele ratio in cDNA was significantly higher than the theoretical ratio of 1 as detected in gDNAs ( P = 0 . 02 , Figure S4 ) , indicating a higher expression of the G- than the C-allele containing TLR7 transcripts in heterozygous PBMCs . The allelic specific expression analysis in EA was similar to our previous findings in Chinese [5] , and confirmed the result of real-time PCR that the G allele of rs3853839 is associated with increased TLR7 mRNA expression . Then , PBMCs were cultured in the absence or presence of the transcriptional inhibitor actinomycin D ( ActD ) , and the G/C allele ratio in cDNA ( normalized to that measured in gDNA ) was determined after 0 , 2 , 4 , 6 , and 24 hours , respectively . As shown in Figure 2D and 2E , the G/C ratio in cDNA appeared to change over time when PBMCs were incubated with ActD and exhibited a statistical difference at the 4 hour point ( P = 0 . 04 ) , implicating slower degradation of the G allele- than the C allele-containing TLR7 transcript in heterozygous PBMCs . The inhibitory effect of ActD on RNA synthesis was corroborated by a decrease in total TLR7 mRNA level at increasing time points after the addition of ActD in PBMC aliquots measured by real-time PCR ( Figure S5 ) . MicroRNAs ( miRNAs ) that bind to target sequences located within the 3′UTR of mRNAs by base pairing have been shown to result in accelerated mRNA turnover or translation repression [6] . Single nucleotide change either within or around the sequence of miRNA target sites can potentially alter the base-pairing patterns and affect miRNA-mediated regulation [7] , [8] . The updated TargetScan database ( Release 6 . 2; http://www . targetscan . org ) indicates that rs3853839 is located within a binding site of miR-3148 , where the non-risk allele ( C ) , but not the risk allele ( G ) , is predicted to match miR-3148 at the second position ( Figure 3A ) . We hypothesized that the C to G variation of rs3853839 could reduce the binding and regulation incurred by miR-3148 , therefore , leading to dysregulated TLR7 expression . We first showed that transcript levels of miR-3148 and TLR7 were inversely correlated in PBMCs from 16 patients with SLE and 21 healthy controls ( R2 = 0 . 255 , P = 0 . 001; Figure 3B ) , suggesting the possible regulation of TLR7 expression by miR-3148 . Next , to verify whether allelic variation of rs3853839 affects the interaction of miR-3148 with TLR7 3′UTR , psiCHECK-2 vectors containing TLR7 3′UTR segment with either the C or G allele of rs3853839 were cotransfected with various doses of miR-3148 or nontarget control mimic into HEK-293 cells . As shown in Figure 3C , we observed significant dose-dependent miR-3148-mediated decrease in luciferase activity for the C-allele construct ( P = 0 . 0003 over all miR-3148-treated C-allele vector groups , ANOVA test ) , but not for the G-allele construct ( P = 0 . 14 ) . Cotransfection with miR-3148 at a concentration of 6 , 12 , and 48 nM , respectively , led to greater than two-fold reduction of luciferase activity in the C-allele than the G-allele construct [reduction in C-allele vs . G-allele construct: 13 . 2% vs . 4 . 8% , P = 0 . 023 ( 6 nM ) ; 22 . 5% vs . 9 . 9% , P = 0 . 0012 ( 12 nM ) ; 21 . 4% vs . 8 . 5% , P = 0 . 0031 ( 48 nM ) ] . These data supported the bioinformatic prediction that miR-3148 directly targets TLR7 3′UTR and the C to G variation of rs3853839 within the binding site alters the inhibitory effect of miR-3148 on modulating TLR7 expression .
Fine-mapping of the TLR7-TLR8 region with high-density genetic markers based on large scale genotyping and imputation confirmed SNP rs3853839 at TLR7 3′UTR as the most likely causal variant responsible for the association of TLR7-TLR8 region with SLE in populations of EA , AA and HS ancestry . In accordance with our previous observation in Asians [5] , we detected elevated TLR7 expression at both mRNA and protein levels in PBMCs from EA homozygous risk G allele carriers , as well as a higher level of the risk than the non-risk allele-containing TLR7 transcripts in EA heterozygous PBMCs . The fact that two distinct ancestries share the same genotype-phenotype association implicates an important regulatory effect of rs3853839 on TLR7 expression . Toward this end , we have extended functional studies showing slower degradation of the risk allele-containing TLR7 transcripts in heterozygous PBMCs and regulation of TLR7 expression by miRNA-3148 that targets 3′UTR at the position of rs3853839 . Finally , we showed that the presence of the risk G allele resulted in reduced suppression by miRNA-3148 , suggesting a likely mechanism for increased TLR7 expression in risk-allele carriers . The importance of TLR7 upregulation on mediating autoimmune responses has been addressed in murine models of SLE . The Y-linked autoimmune accelerator ( Yaa ) modifier , suggested mainly due to Tlr7 gene duplication , provides a prime example of TLR7 dysregulation leading to autoreactivity and inflammatory pathology [9]–[11] . Increasing Tlr7 gene dosage via generation of transgenic mice results in development of systemic autoimmunity , the severity of which directly correlates with the degree of Tlr7 overexpression [12] . Increased Tlr7 gene dosage promotes autoreactive lymphocytes activation , dendritic cells proliferation , and secretion of proinflammatory cytokines and IFN-α [12] , which in turn upregulates TLR7 expression , leading to a feedback loop exacerbating autoimmunity [13] . In patients affected with SLE , up-regulated expression of TLR7 mRNA has been reported in PBMCs and B cells [14] , [15] . Although a copy number variation ( CNV ) study in Mexican population showed increased TLR7 copies in childhood-onset SLE patients [16] , no evidence for common CNVs at the TLR7-TLR8 region has been identified in individuals of diverse ancestries through our previous study by three independent methods including quantitative real-time PCR , PmeI pulsed-field gel electrophoresis and Southern blot [5] , two recent studies using customized CGH platforms [17] , [18] as well as other studies listed in the Database of Genomic Variants ( http://projects . tcag . ca/variation; the latest version released in November 2010 ) , suggesting that mutations similar to Yaa are not a frequent feature of human SLE . The current study identifying genetic variations conferred by a regulatory SNP in TLR7 expression and SLE susceptibility suggests that murine models provide profound clues to human genetics if we look beyond the specific mutations identified in the relevant pathways . Unlike our findings in Asians that both sexes showed association [5] , the impact of rs3853839 on risk for SLE was only observed for women in the non-Asian datasets ( Table 1 ) . Given the low prevalence of SLE in men , it is often challenging to collect a large enough number of affected men in a given population . Under the assumption that the associated G allele confers genetic risk with an odds ratio of 1 . 26 in EA , 1 . 22 in AA and 1 . 29 in HS subjects ( ORs were determined in female datasets ) , and considering P<0 . 05 as the threshold of significance , the power estimate for female samples in each ancestry reaches more than 85% , whereas for male samples it is only 50% in EA , 19% in AA and 25% in HS dataset . Thus , there was clearly inadequate power to evaluate this association in AA and HS men . Despite a relatively robust sample size of EA men ( 344 SLE vs . 1 , 151 controls ) , a significantly higher G allele frequency was observed in male than female controls ( 20 . 0% vs . 16 . 5% , P = 0 . 005 ) , contributing to the difficulty in assessing association with SLE in EA male subjects . To our knowledge , the association of rs3853839 ( or its tag SNP ) with SLE has not been reported in four SLE GWA studies in European-derived populations [19]–[22] and three GWA studies in Asians [23]–[25] . According to the 1000 Genomes Project data , rs3853839 locates in a region with poor LD structure and cannot be tagged by any known SNP at the TLR7-TLR8 region with r2>0 . 65 . The SNP rs850632 , located at TLR7 3′downstream , shows the strongest LD with rs3853839 in Europeans ( r2 = 0 . 38 ) and Asians ( r2 = 0 . 65 ) . However , neither rs3853839 nor rs850632 has been included in predesigned commercial genotyping arrays of those GWA studies , resulting in the absence of associations . Even if rs3853839 was genotyped , the published GWA studies might have inadequate statistical power to capture its association in the initial discovery analyses [5] . Evidence of other TLR7 polymorphisms associated with SLE has been reported , including two intronic SNPs ( rs179019 and rs179010 ) found in Japanese population [26] and an exonic SNP ( rs179008 ) in individuals from Southern Brazil [27] . The reported associations were modest due to limited sample size of these studies ( less than 400 cases and 450 controls ) , and none of them have been confirmed by the current fine-mapping study using a large collection of EA , AA and HS cases-controls ( Table S1 ) . TLR8 polymorphisms have been described in infectious diseases [28] , [29] with a genetic effect localized to a functional variant at exon 1 ( rs3764880 , Met1Val ) . The G allele of rs3764880 , which abolishes a putative start codon within the alternative TLR8 transcript isoform a ( Figure 1A ) , conferred a protective effect on susceptibility to pulmonary tuberculosis in Indonesian and Russian men [28] , as well as on HIV disease progression in Germans [29] . Our data showed a significantly increased frequency of rs3764880-G allele in SLE than healthy controls in the three non-Asian datasets; however , its association with SLE was dependent on that of TLR7 SNP rs3853839 . Other variants at the TLR7-TLR8 region showed either weaker association than rs3853839 in trans-ancestral meta-analysis or association uniquely in EA or HS . Taken together , these data support rs3853839 as the most likely polymorphism associated with SLE shared by multiple ancestries . Although imputation facilitated our ability to capture common variants ( MAF>1% ) , further refinement in genetic effects of rare variants ( MAF<1% ) is needed by deep sequencing of this locus , especially the intergenic region between TLR7 and TLR8 that was not well imputed in this study . Variations in 3′UTR regions may be important in gene regulation . To date , expression quantitative trait loci ( eQTL ) mapping has been widely used for characterization of SNPs that affect gene expression [30] . Although the TLR7 expression has been measured in previous whole-genome eQTL studies , currently only those using EBV-transformed lymphoblastoid cell lines of 1000 Genomes Project individuals provide publically available genotyping data of rs3853839 . Based on the study by Stranger et al [31] , we found that CG carriers of rs3853839 showed elevated TLR7 expression compared with CC carriers in YRI women ( P = 0 . 012 ) . In male individuals , the G allele of rs3853839 showed a trend of association with elevated TLR7 expression in CHB+JPT , CEU and YRI men , and the association was significant when combining all male data ( P = 0 . 014 ) . These findings are consistent with our results that rs3853839 alleles are associated with differential TLR7 expression . An important finding of this study is that the SLE-associated variant rs3853839 confers a genetic effect on modulation of TLR7 expression by an epigenetic factor miR-3148 . Accumulating evidence suggests that miRNAs are fine tuners of TLR signaling pathways [32] . Regulation by miRNA may occur at various levels of TLR pathways by targeting adaptor molecules , downstream regulators and cytokines ( reviewed in [32] , [33] ) . However , few studies point to TLR themselves ( e . g . TLR2 and TLR4 ) being directly targeted by miRNAs [34] , [35] . Using algorithms from TargetScan , only the newly identified human miR-3148 [36] , which is not evolutionarily conserved among mammals , is predicted to bind TLR7 3′UTR sequences at the position of rs3853839 . The inverse correlation of miR-3148 and TLR7 levels in PBMCs , along with functional validation by reporter gene assay , confirms an inhibitory effect of miR-3148 on regulating TLR7 expression and allelic variation of rs3853839 affecting miRNA-mRNA interactions . Further study will focus on investigating miR-3148 expression patterns in specific immune cell types , assessing biological impacts of changes in miR3148-mediated TLR7 expression on downstream immune responses , and evaluating roles of other miRNAs that target sequences in the vicinity of rs3853839 . Of interest , an unconventional role for miRNAs has been identified as endogenous activators for RNA-sensing receptors ( TLR7/8 ) in a cell- or tissue-type specific manner [37] , [38] . Therefore , miRNA regulation in TLR7 signaling is more complicated than we expected and further functional studies showing the exact effects of miRNAs on TLR7 responses are warranted . In summary , we have advanced our previous study by showing rs3853839 ( at TLR7 3′UTR ) as the most likely polymorphism responsible for the association of TLR7-TLR8 region with SLE in individuals of EA , AA and HS ancestry , and have characterized a differential miR-3148 modulation which explains the effect of allelic variation of rs3853839 on TLR7 expression . Our study highlights the importance of TLR7 as a shared genetic contributor to SLE in multiple ancestries , and provides evidence that microRNA acts as a negative regulator to control TLR7 expression , suggesting the possibility of miRNA-based therapies for amelioration of autoimmune diseases such as SLE where excessive TLR7 activation exists .
Written informed consent was obtained from all study participants and each participating institution had Institutional Review Board ( IRB ) approval to recruit samples . The overall study was approved by the IRB of the Oklahoma Medical Research Foundation ( OMRF ) . To test the association of TLR7-TLR8 with SLE , we used a large collection of case-control subjects from the collaborative Large Lupus Association Study 2 ( LLAS2 ) , including European American ( 4 , 248 cases vs . 3 , 818 controls ) , African American ( 1 , 724 cases vs . 2 , 024 controls ) , and Hispanic enriched for the Amerindian-European admixture ( 1 , 622 cases vs . 887 controls ) . African Americans included 286 Gullahs ( 155 cases vs . 131 controls ) , who are subjects with African ancestry . Cases were defined by meeting at least four of the 1997 American College of Rheumatology ( ACR ) revised criteria for the classification of SLE [39] . DNA samples were processed at the Lupus Genetics Studies Unit of OMRF . SNP genotyping was performed using an Illumina custom bead array on the iSCAN instrument for 47 SNPs covering the TLR7-TLR8 region on Xp22 . 2 and 347 admixture informative markers ( AIMs ) . SNPs meeting the following criteria were included in the association analysis: well-defined cluster scatter plots , SNP call rate >90% , minor allele frequency >1% , total proportion missing <5% , P>0 . 05 for differential missing rate between cases and controls , and Hardy-Weinberg proportion ( HWP ) test with a P>0 . 01 in controls and P>0 . 0001 in cases . Subjects with genotype missing rate >10% ( due to low quality ) , shared identical by descent >0 . 4 or showing mismatch between the reported and estimated gender were removed . The global ancestry of each subject was estimated based on genotype of AIMs using principal components analysis [40] and ADMIXMAP [41] , as described in another LLAS2 study [42] , and then genetic outliers were removed . Finally , a total of 13 , 339 unrelated subjects , including European Americans ( EA: 3 , 936 cases vs . 3 , 491 controls ) , African Americans ( AA: 1 , 679 vs . 1 , 934; composed of 92 . 5% of African Americans and 7 . 5% Gullahs ) and Hispanics enriched for the Amerindian-European admixture ( HS: 1 , 492 vs . 807 ) , were analyzed for 41 genotyped SNPs of TLR7-TLR8 . Imputation was performed at 12 . 86–12 . 95 Mb on Xp22 . 2 using IMPUTE 2 . 1 . 2 [43] , with SNP/INDEL genotypes of 381 Europeans , 246 Africans and 181 Americans from the 1000 Genomes Project ( “version 3” of the Phase 1 integrated data , March 2012 release ) as references in imputation for our EA , AA and HS subjects , respectively . Imputed genotypes had to meet information score of >0 . 9 , as well as the quality control criteria as described above . After imputation , we obtained an additional 75 variants for EA , 57 for AA and 63 for HS ( the number varied based on LD structure ) for further analysis . Total RNA was purified with TRIzol reagent ( Invitrogen ) from PBMCs and reverse-transcribed into cDNA with Superscript II Reverse Transcription kit ( Invitrogen ) . The mRNA levels of TLR7 ( NM016562 . 3 ) and TLR8 ( isoform a: NM138636 . 4 and isoform b: AF246971 . 1 ) were measured by quantitative real-time PCR using TaqMan assays ( TLR7 probe: Hs00152971_m1; TLR8 isoform a probe: Hs00607866_mH; TLR8 isoform b probe: Hs00152972_ml , Applied Biosystems ) . All samples were run in triplicate . Relative expression levels of TLR7 and TLR8 were normalized to the level of RPLP0 , calculated by the 2−ΔΔCt method and Log10 transformed . The association of rs3853839 with mRNA levels of TLR7 or TLR8 was evaluated using ANOVA , Student's t and linear regression test . To examine the correlation of miR-3148 and TLR7 mRNA levels , total RNA enriched in small RNAs were isolated from PBMCs using mirVanaTM miRNA isolation kit ( Invitrogen ) , followed by reverse transcription with TaqMan MicroRNA Reverse Transcription kit ( Applied Biosystems; for detecting miR-3148 ) and Superscript II Reverse Transcription kit ( Invitrogen; for detecting TLR7 ) , respectively . The miR-3148 level was quantified using Taqman MicroRNA Expression assay ( Applied Biosystems ) , and the TLR7 level was measured using the same probe as described above . All samples were run in triplicate . Relative expression levels of miR-3148 and TLR7 were normalized to the level of snRNA U6 and RPLP0 , respectively , calculated by the 2−ΔΔCt method and Log10 transformed . Association between transcript levels of TLR7 and miR-3148 was evaluated using linear regression test . Four-color flow cytometry was performed to investigate intracellular expression of TLR7 and TLR8 in PBMCs from healthy EA and Asian individuals who were homozygous for rs3853839 ( 7 pairs of G-allele vs . C-allele carriers in each gender group ) . Freshly isolated PBMCs were incubated with 2% pooled human serum to block nonspecific binding to Fcγ receptors and then incubated with peridinin chlorophyll protein ( PerCP ) -conjugated anti-human CD3 , allophycocyanin ( APC ) -conjugated anti-human CD19 and phycoerythrin ( PE ) -conjugated or fluorescein isothiocyanate ( FITC ) -conjugated anti-human CD14 ( Miltenyi Biotec ) to identify T cell , B cell and monocyte subpopulations , respectively . For intracellular staining , PBMCs were fixed in Fixation buffer ( R&D Systems ) for 10 minutes at room temperature , washed twice in Permeabilization/Wash buffer ( R&D Systems ) and stained with PE-conjugated mouse anti-human TLR7 mAb ( R&D Systems ) and FITC-conjugated mouse anti-human TLR8 mAb ( Imgenex ) for 1 hour at room temperature . Background fluorescence was assessed using appropriate isotype- and fluorochrome-matched control antibodies . Cells were collected and analyzed by FACSCalibur flow cytometer equipped with the manufacturer's software ( CellQuest; BD Biosciences ) . Student's t test was used to compare protein levels of TLR7 or TLR8 in PBMCs from individuals of different genotypes . The fragment of TLR7 3′-UTR bearing the G or C allele of rs3853839 was amplified by PCR from genomic DNA of subjects homozygous for the G or C allele using the following primers: 5′-TGTCTCGAGCCCTTCTTTGCAAAAC-3′ ( forward ) and 5′-AGAGCGGCCGCTAGTTGGCTCCAGCAAT-3′ ( reverse ) . The PCR products were inserted into the downstream of the Renilla luciferase gene in the reporter vector psiCHECK-2 ( Promega ) by digestion using the restriction enzymes Not I and Xho I . The psiCHECK-2 vector also contained a firefly luciferase gene to serve as an internal transfection normalization control . All constructs were sequenced to assure proper orientation and authenticity in the vector . HEK-293 ( human embryonic kidney cell line ) and HL-60 ( human leukemic cell line ) cells were obtained from the American Type Culture Collection ( ATCC ) . HEK-293 cells were maintained in Dulbecco's modified Eagle's medium supplemented with 10% FBS , seeded on a 24-well plate at a concentration of 2×105 cells/well , and transiently transfected using Lipofectamine 2000 ( Invitrogen ) with 1 µg of either rs3853839 G or C reporter construct . HL-60 cells are predominantly a neutrophilic promyelocyte ( precursor ) and can be induced to differentiate to neutrophil-like cells when grown in RPMI 1640 medium with 15% FBS plus 2 mM L-glutamine , 25 mM HEPES and 1 . 25% DMSO [44] . Differentiated HL-60 cells seeded on 24-well plates ( 2×106 cells/well ) were electroporated with 3 µg of report construct on a nucleofector device ( Amaxa ) . The luciferase activity in total cell lysates was measured after 24 hours using a dual luciferase reporter assay system ( Promega ) . Renilla luciferase activities were normalized to firefly luciferase activities . Each transfection was performed in quadruplicates and triplicates for HEK-293 and HL-60 cells respectively , and luciferase assays were repeated four times . MicroRNA hsa-miR-3148 and nontarget control ( NC ) mimics were synthesized by Thermo Fisher Scientific . To test the effect of miR-3148 , HEK-293 cells plated in 96-well plates were transiently cotransfected with 100 ng of each reporter construct ( psiCHECK-2 empty vector , rs3853839-G or -C allele constructs ) and increasing concentrations ( 1 , 6 , 12 and 48 nM ) of miR-3148 or nontarget control mimic using Lipofectamine 2000 reagent ( Invitrogen ) , and luminescence was measured 24 hours after transfection . Each transfection was performed in quadruplicates and repeated three times . Luciferase activity of reporter vectors was compared using Student's t test . PBMCs isolated from EA healthy women with the GC genotype of rs3853839 ( n = 7 ) were cultured in the absence or presence of 5 µg/mL ActD for 0 , 2 , 4 , 6 and 24 hours . Using real-time PCR , we detected a decrease in total TLR7 mRNA levels over time with ActD incubation , which confirmed the transcriptional inhibition by ActD and allowed for detection of allelic differences in mRNA degradation . The G/C allelic ratio in the cDNA and gDNA after treatment of PBMCs with or without ActD were determined by pyrosequencing and calculated using software PSQMA 2 . 1 ( Biotage ) as previously described [5] . The G/C allele ratio obtained in TLR7 transcripts was normalized to that measured from gDNA of the same sample . A paired t test was used to compare the mean G/C allele ratio in TLR7 transcripts in PBMCs treated with ActD or vehicle control at each time point . Associations of SNPs with SLE were assessed in each ancestral group under a logistic regression model adjusted for gender and the first three principal components estimated using AIMs . Conditional haplotype-based association tests were also performed by adjusting for gender and the first three principal components . The trans-ancestral meta-analysis was conducted on 40 genotyped and 14 imputed SNPs that were shared by the three ancestries with both a fixed and random-effects model . Homogeneity of odds ratios was evaluated using Cochrane's Q test . For each SNP , if the Cochran's Q test showed no evidence of genetic heterogeneity ( P>0 . 05 ) , a fixed-effects model was implemented; otherwise , a random-effects model was used . The Bonferroni corrected P-value threshold was adjusted to P<9 . 1×10−4 on the basis of the maximum number of tests across all populations ( 55 independent variants with r2<0 . 8 ) . All analyses described above were performed using PLINK v1 . 07 . Pairwised LD values shown in Figure 1 and Figure S1 were calculated using Haploview 4 . 2 . Other data were analyzed using GraphPad Prism 4 . 0 software . A P value<0 . 05 was considered to be statistically significant . | Systemic lupus erythematosus ( SLE ) is a debilitating autoimmune disease contributed to by excessive innate immune activation involving toll-like receptors ( TLRs , particularly TLR7/8/9 ) and type I interferon ( IFN ) signaling pathways . TLR7 responds against RNA–containing nuclear antigens and activates IFN-α pathway , playing a pivotal role in the development of SLE . While a genomic duplication of Tlr7 promotes lupus-like disease in the Y-linked autoimmune accelerator ( Yaa ) murine model , the lack of common copy number variations at TLR7 in humans led us to identify a functional single nucleotide polymorphism ( SNP ) , rs3853839 at 3′ UTR of the TLR7 gene , associated with SLE susceptibility in Eastern Asians . In this study , we fine-mapped the TLR7-TLR8 region and confirmed rs3853839 exhibiting the strongest association with SLE in European Americans , African Americans , and Amerindian/Hispanics . Individuals carrying the risk G allele of rs3853839 exhibited increased TLR7 expression at the both mRNA and protein level and decreased transcript degradation . MicroRNA-3148 ( miR-3148 ) downregulated the expression of non-risk allele ( C ) containing transcripts preferentially , suggesting a likely mechanism for increased TLR7 levels in risk-allele carriers . This trans-ancestral mapping provides evidence for the global association with SLE risk at rs3853839 , which resides in a microRNA–gene regulatory site affecting TLR7 expression . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"systemic",
"lupus",
"erythematosus",
"medicine",
"rheumatology",
"immunity",
"genetic",
"association",
"studies",
"innate",
"immunity",
"genetics",
"immunology",
"biology",
"human",
"genetics",
"genetics",
"of",
"disease",
"genetics",
"and",
"genomics"
] | 2013 | MicroRNA-3148 Modulates Allelic Expression of Toll-Like Receptor 7 Variant Associated with Systemic Lupus Erythematosus |
Spatial structure and local migration are predicted to promote the evolution of less aggressive host exploitation strategies in horizontally transmitted pathogens . Here we explore the effect of spatial structure on the evolution of pathogens that can use both horizontal and vertical routes of transmission . First , we analyse theoretically how vertical transmission can alter evolutionary trajectories and confirm that space can impede the spread of virulent pathogens . Second , we test this prediction using the latent phage λ which transmits horizontally and vertically in Escherichia coli populations . We show that the latent phage λ wins competition against the virulent mutant λcI857 in spatially structured epidemics , but loses when spatial structure is eroded . The vertical transmission of phage λ immunizes its local host pool against superinfection and prevents the spread of the virulent λcI857 . This effect breaks down when mixing facilitates horizontal transmission to uninfected hosts . We thus confirm the importance of spatial structure for the evolutionary maintenance of prudent infection strategies in latent viruses .
When individuals compete for a common resource natural selection often favours more aggressive exploitation strategies . This may lead to resource exhaustion and , consequently , to population extinction , a process known as the 'tragedy of the commons'[1] . Because pathogens compete for a common resource ( the host population ) , the same process may select for extreme exploitation strategies . Yet , two main ecological factors can alter this evolutionary outcome and promote the evolution of less aggressive exploitation strategies . First , epidemic spread reduces the density of susceptible hosts and can feed back on the selective pressure that favours intermediate exploitation strategies [2–5] . Indeed , decreased availability of susceptible hosts weakens selection for higher transmission rates [6–8] . Ultimately , this may select for intermediate evolutionarily stable strategies , balancing the benefit ( transmission ) and the cost ( virulence: induced host mortality ) of host exploitation . We recently tested this idea using the bacteriophage λ in experimental epidemics spreading through well-mixed environments [9] . We confirmed that more virulent strains are indeed selected for during early epidemics , when uninfected hosts are abundant , but also that natural selection favours latent strains of the virus as disease prevalence increases . Second , theoretical studies on spatially structured populations indicate that localized transmission often favours more prudent host exploitation strategies [10–21] . Indeed , when epidemics are spatially structured , extreme strategies of host exploitation may lead to the over-exploitation of the local host supply and therefore fail to invade . The intensity of local competition for susceptible hosts depends , however , on the precise genetic and epidemiological structure of the pathogen population [21] . The effect of spatial structure on the evolution of pathogens has had some experimental support [22–23] which is , however , based on experiments with obligate killing and strictly horizontally transmitting pathogens . Previous experimental work has shown that vertical transmission selects for lower virulence [24–25] but the effect of spatial structure on vertically transmitted pathogens remains an open question . In the present study we explore the effect of spatial structure on the evolution of pathogens that can be transmitted both vertically and horizontally . We first develop and analyse a theoretical model to understand the evolutionary dynamics of pathogens in spatially structured epidemics ( Fig 1 ) . This model allows infected hosts to reproduce and to transmit their pathogens vertically . We use this model to study the competition between different pathogen genotypes under various levels of mixing ( see supporting information in S1 Text ) . Then , we confront these predictions with experiments performed with the horizontally and vertically transmitting temperate bacteriophage λ . Bacteriophage λ is an avirulent pathogen that controls its own replication by the virulence repressor cI and integrates into the host genome as a prophage without killing the host cell—a process known as lysogenization . The virulence repressor cI furthermore controls prophage λ reactivation and excludes superinfection by a second viral particle [26] . Yet , spontaneous mutations in cI can markedly increase virulence [27] . We studied competition between temperate bacteriophage λ and its virulent mutant λcI857 . Mutant λcI857 carries an unstable virulence repressor protein ( cI857 ) that biases the viral life cycle towards horizontal transmission: whereas λ lysogenizes about 60% of newly infected cells killing only about 40% of cell , λcI857 kills 96% of host cells upon first infection ( see Fig S2B in S1 Text ) . In addition λcI857 shows a higher rate of prophage reactivation , sacrificing infected host cells for increased production of free viral particles ( Fig S2A in S1 Text ) . Furthermore , λcI857 is less efficient in the exclusion of superinfection ( see Fig S2B in S1 Text ) . Whereas the temperate λ wildtype shows an avirulent host exploitation strategy the virulent mutant λcI857 readily exploits and kills its host cell . To experimentally investigate the effect of spatial structure on the success of these opposing strategies we competed λ and λcI857 in experimental epidemics with different degrees of spatial structure ( Manipulation of spatial structure is described in Material & Methods and [28] ) .
We study a spatial version of the compartmental model depicted in Fig 1 . This allows us to make predictions on the epidemiological and evolutionary consequences of spatial structure . First , Fig 2A shows that lower levels of mixing slow down the spread of the infection because in a spatially structured environment , pathogens experience a lower density of susceptible hosts . Second , the change in frequency of a virulent mutant strain is ( see Supporting Information Theory in S1 Text ) : dfdt=f ( 1−f ) [βm[S|Im]−βw[S|Iw]︸Horizontaltransmission+δbI ( [o|Im]−[o|Iw] ) ︸Verticaltransmission− ( αm−αw ) ︸Virulence] ( 1 ) The first term measures the contribution of the horizontal transmission route and depends on the local density of susceptible hosts in the neighbourhood of a host infected by a mutant or wild-type parasite The second term measures the contribution of vertical transmission and depends on the local density of empty sites experienced by hosts infected by each type of parasite . Indeed , a vertical transmission event is conditional upon a reproduction event and thus on the availability of empty sites in the neighbourhood . Finally , the last term measures the cost of virulence . Fig 2B shows numerical simulations based on the pair approximation [29–31] under different levels of mixing ( see supporting information Theory in S1 Text ) . Note that , under all level of mixing , the frequency of the virulent mutant initially increases . Indeed , early on during the epidemics , parasites have access to a large density of susceptible hosts , which favours the spread of more transmissible mutants . This benefit , however , is only transient . As the pool of susceptible hosts is exhausted , more prudent pathogens end up being selected under all mixing treatments . Spatial structure , however , has a dramatic impact on the virulent mutant: less mixing drastically reduces the transient benefit of virulence . To understand this effect , we represent on Fig 2C the effect of spatial structure on the different components of the selection coefficient in Eq ( 1 ) . Spatial structure acts through two different effects: ( i ) space reduces the horizontal transmission benefit ( red curve in Fig 2C ) and ( ii ) space generates benefit for the virulence mutant through vertical transmission ( blue curve in Fig 2C ) . The first effect is the classical effect of space often discussed as a self-shading or kin selection argument [14 , 21 , 32] . The second effect is due to the accumulation of empty sites near virulence mutants , which is caused by the death of hosts infected by related ( and thus virulent ) genotypes . In other words , virulence mutant have more opportunities to transmit vertically than avirulent genotypes . In addition , to better grasp the effect of space on pathogen evolution , we track the change in mutant frequency over two types of infected hosts: hosts that acquired the infection horizontally and hosts that acquired the infection vertically . Fig 2B reveals the effect of mixing on the change in frequency of the mutant in these two compartments . In a well-mixed environment the frequency of the mutant is always higher in the horizontally infected hosts . In contrast , in a structured environment , the frequency of the virulent mutant can be higher in vertically infected hosts . This is due to the effect of spatial structure on the opportunities for vertical transmission . As explained above the virulent mutant has more opportunities for vertical transmission because the local density of empty sites is higher ( see Fig 2B ) . Another way to see this effect is to track the ratio WH / WV which provides a measure of the relative contribution of the two routes of transmission to mutant fitness . When this ratio is above 1 the change in mutant frequency is mainly governed by the horizontal transmission route . In contrast , when it is below one the vertical transmission route contributes more to the mutant fitness . Fig 3 shows that during the early stage of the epidemics horizontal transmission drives the evolution of the virulent mutant under all mixing treatments but that the contribution of vertical transmission is always increased by spatial structure . The above model is an attempt to capture the interplay between the effects of spatial structure and vertical transmission on the evolution of pathogen virulence . To test these predictions experimentally we followed competition of the avirulent λ and virulent λcI857 in spatial epidemics after an initial inoculation of a 1:1 ratio of cells infected by λ or λcI857 onto a biofilm of susceptible cells . Thereupon spatial structure was either left undisturbed or disturbed for 30s or 24h in the absence of liquid , or 24h in the presence of liquid ( 24h-wet ) . By engineering fluorescent protein expression cassettes ( CFP and YFP ) into the viral genome we tagged infected cells by a fluorescent colour . This enabled us to directly visualize pathogen-host structure and spatial prevalence in these four environments ( see Fig 4 ) . In the undisturbed treatment the epidemic spreads in a circular front segregating by viral types ( λCFP or λYFP ) but large areas remain uninfected ( see Fig 4A ) . In the 30s disturbance treatment the epidemic spreads through the entire biofilm and creates epidemic clusters dominated by a single type of virus ( Fig 4B ) . After 24h disturbance epidemic clusters disappear , but the biofilm remains structured ( Fig 4C ) . In the 24h-wet treatment epidemic structure is completely homogenized ( Fig 4D ) . Thus , as expected from our model simulation ( Fig 2A ) , higher mixing is associated with a significant increase in overall prevalence of infection ( χ12 = 36 . 52 , p<0 . 0001 , Fig S3 in S1 Text ) . The mixing treatment strongly affects competitive fitness W of λcI857 ( χ12 = 79 . 15 , p<0 . 0001 , Fig 5 ) . In the undisturbed environment λcI857 fitness decreases by ~100 fold after the first transfer day , in contrast in the 24h-wet environment λcI857 fitness increases by ~5 fold . As expected , intermediate spatial disturbance treatments ( 30s and 24h ) yield intermediate levels of λcI857 fitness ( Fig 5 ) . This demonstrates that erosion of the spatial structure increases the fitness of the virulent λcI857 by ~500 fold . Note that additional transfers did not affect this pattern except for the treatment 30s where λcI857 fitness increases with time ( t = 5 . 08 , p<10–4 ) . To disentangle the contribution of horizontal and vertical transmission to the fitness of λcI857 we repeated the experiment for the first round of spatial transfer with uninfected cells marked by the red fluorescent protein mCherry . This way we identified new infections from horizontal transmission as doubly coloured cells ( mCherry and colour of the new infecting virus ) and vertical transmission from the initial lysogen as singly coloured cells ( colour of the originally integrated virus ) . From these data we calculated the competitive fitness of λcI857 among all new infections ( horizontal fitness WH ) and fitness among all vertical transmission events ( vertical fitness WV ) in each spatial environment . Fig 6 presents the effect of mixing on the ratio WH / WV between horizontal and vertical fitness components of λcI857 . In the absence of mixing both routes of transmission contribute to λcI857 fitness . Interestingly , in agreement with our theoretical prediction ( Fig 2C ) , we found that higher levels of mixing promote horizontal transmission ( χ12 = 14 , 82 , p = 0 . 002 ) . For instance , in the highly mixed environment ( 24h-wet treatment ) , 3 out of 4 new infections result from the horizontal transmission route ( Fig 6 ) .
During an epidemic , disease prevalence increases , and the opportunities to find uninfected hosts dwindle . This may select for prudent exploitation of the remaining host population . In a previous study , we demonstrated the impact of such epidemiological feedback on the evolution of bacteriophage λ [9] . Here , we use the same experimental system to study the effect of spatial structure on the evolution of bacteriophage λ . To account for latency and vertical transmission in the life cycle of bacteriophage λ , we extend previous theory on the effect of spatial structure on pathogen evolution by allowing for vertical transmission of the pathogen . Our model tries to capture all the details of the bacteriophage λ life-cycle , but for tractability we do not explicitly model the free living stage of the virus . Our central result is that the fitness of a virulent strain relative to a more prudent strain decreases as the environment becomes increasingly structured . A general insight of previous theory is that the outcome of selection on parasite life-history traits depends on the balance between the genetic and epidemiological structures of the parasite population [21] . At the genetic level , parasites infecting a cluster of hosts tend to be related when transmission occurs locally . Hence , competition for the local host supply tends to take place among related pathogens ( kin competition ) [18 , 21 , 32] . At the epidemiological level , local pathogen spread reduces the local density of uninfected hosts , thereby degrading the parasite's local environment and diminishing future opportunities for horizontal transmission [5 , 14] . From models without vertical transmission , it has been shown that more prudent strategies of host exploitation are selected for if the genetic structure of the parasite population is above a threshold determined by the epidemiological structure of the host population [33] . As interactions become more localized , genetic structure increases and the balance is tilted at the advantage of more prudent strains [21] . We show that vertical transmission does not alter the general theoretical expectation that spatial structure should select for more prudent host exploitation strategies . Yet our analysis reveals that in a spatially structured environment vertical transmission can affect evolutionary dynamics via two main effects . First , vertical transmission affects opportunities of horizontal transmission through a modification of the local density of susceptible hosts ( the first term in Eq ( 1 ) ) . Second , vertical transmission generates a force ( the second term in Eq ( 1 ) ) that may counteract the effect of spatial structure . During the early stage of the epidemic this effect remains very low but further work is required to investigate the impact of the effects of vertical transmission on evolutionary stable virulence strategies . Our model is useful to decouple the change in mutant frequency in horizontally and vertically infected hosts . This distinction between different host types is artificial but it helps to grasp the interplay between spatial structure and transmission routes . As in other situations where the pathogen can appear in different states [9] or in different hosts [8 , 34–36] , tracking the change in frequency between different compartments enables us to generate a prediction regarding the effect of spatial structure on the relative contribution of the two transmission routes on pathogen fitness ( Fig 2C ) . This new theoretical model generates predictions that we tested experimentally with bacteriophage λ . Unlike our previous experimental study in chemostats [9] , the present experiment has been realized on agar plates to manipulate the amount of spatial structuring . Yet , on a normal agar plate bacterial growth eventually enters stationary phase and the epidemiological and evolutionary dynamics halt . Consequently , our experiments allow us to test predictions during the initial phase of the epidemics only . During this phase we made three qualitative predictions that we verified experimentally . First , we confirmed that lower levels of mixing reduce the speed of the epidemic . Second , we confirmed that a virulent strain is outcompeted by a more prudent strain when the environment becomes increasingly structured . This is our main result and it agrees well with previous experimental studies realized with horizontally transmitted pathogens [22–23] . Third , our experimental setup allows monitoring the relative contribution of horizontal and vertical transmission among new infections and confirms that mixing decreases the contribution of vertical transmission to pathogen fitness . The evolution of host exploitation strategies can often be explained at the level of within-host selection . For parasites competing for within-host resources , more prudent host exploitation should be selected for when the probability of multiple infections decreases [4] . In our model , superinfection exclusion by a latent λ prophage guarantees that only related pathogen individuals compete within the same host cell and , hence , limits selection for increased replication and virulence at the within-host level . Similarly , many covertly infecting and genome-integrating viral pathogens take within- host competition out of the equation by active exclusion of superinfection . In herpes and retroviruses , for example , superinfection exclusion is well described and might be a key requirement for the maintenance of viral latency [37–39] . Yet some mutations are known to enable λ to break through superinfection exclusion [26] . We did not observe such mutants in our present experiments , but we were able to extend our theory to explore the effect of spatial structure on the epidemiology and evolution of such virulence mutants that can superinfect already infected bacteria ( see supporting information Theory in S1 Text ) . This analysis shows that mixing does affect the epidemiology and evolution in this alternative scenario: more mixing enhances the spread of the virulence mutant . Hence , this analysis suggests that the effect of spatial structure on the transient dynamics we report in the present study is robust to this modification of the life cycle . We believe that similar predictions may be feasible to understand and predict the epidemiology and evolution of a broad range of pathogens with complex transmission modes .
The theoretical analysis is presented in the supporting information ( S1 Text ) . We examine the evolutionary dynamics of a vertically and horizontally transmitted pathogen in a spatially structured environment . The analytical model reveals that the change in frequency can be decomposed in different terms that characterize the local environment of competing variants of the pathogens . These terms can be tracked in numerical simulations ( Fig 2 ) . Further understanding of the evolutionary dynamics can be obtained by focusing on the change in frequency of a mutant pathogen in hosts that have acquired the pathogen horizontally ( IH ) and in hosts that have acquired the pathogen vertically ( IV ) . Host exploitation strategies of λ and λcI857 were quantified by three independent assays . ( 1 ) Virus production ( PFU/mL ) : Lysogens were grown to OD600nm = 0 . 6 at 30°C and shifted to 35°C for 2h until lysis occurred . Viral titers were determined by qPCR ( Roche LightCycler480 primers F:5’AATGAAGGCAGGAAGTA3’ R:5’GCTTTCCATTCCATCGG3’ ) on filtered lysates ( Millipore 96-well filtration plate Multiscreen HTS ) . CP values were calibrated to PFU/ml by top-agar plating a dilution series of a λvir ( 3x109 pfu ) . ( 2 ) Growth of infected hosts ( CFU/mL ) : Lysogens were diluted to OD600nm = 0 . 07 and grown for 6h at 35°C in eight replicates in 96-well plates ( 900 rpm , Titramax shaker ( Heidolph , Germany ) ) . Final OD600nm was measured in an Infinity200 microplate reader ( Tecan , Austria ) . OD600nm values were calibrated to CFU/ml by plating a dilution curve . ( 3 ) Lysogenization rate was determined by challenging non-infected E . coli MG1655 of OD600nm = 0 . 1 with 108 PFU/mL free virus particles of λCFP , λYFP , λcI857CFP and λcI857YFP for 24h . After 24h , the proportion of lysogenized ( fluorescent ) cells was determined by flow cytometry ( Fig S2 in S1 Text ) . Fluorescently labelled phage were constructed as described in [9] . Spatial epidemics were started by spreading a 108 cfu/ml suspension of uninfected E . coli MG1655 ( starved 24h in 0 . 2%Maltose , 10mM MgSO4 ) onto agar plates until dried . Thereafter a 1:1 mixture of cells infected by λ and λcI857 was inoculated in localized spots by an array of 0 . 4mm stainless steel tattooing needles at about 4mm distance . Epidemic structure was disturbed by rolling 4mm sterile stainless steel beads over the dry agar surface for 30s or 24h or over agar surface humidified by sterile saline solution ( 24h-wet ) or kept undisturbed . Agar plates were incubated at 35°C in agitation ( 250rpm ) , but only in the 24h and 24h-wet treatment steel beads remained on the agar surface . To control for marker effects all competitions were carried out with reciprocal marker-mutant combinations ( λYFP vs . λcI857CFP and λCFP vs . λcI857YFP ) . Spatial structure was photographed on an Olympus BH-2 RCF microscope with a 2x fluorite objective and filter cube #69380 ( Chroma Thechnology Corp . , VT , USA ) . For the daily transfer of the epidemic structure the needle array described above was poked into plates that had been incubated for 24 hours and subsequently the needle array was poked into agar plates containing a layer of uninfected cells as described above . For flowcytometry the bacterial layer was washed off with 2 ml saline and analysed on a B&D Fortessa flowcytometer ( Excitation CFP , YFP , mCherry at 405 , 488 and 561 nm ) . Horizontal and vertical transmission was estimated by repeating the spatial epidemics for a single cycle with lysogens of λCFP , λYFP , λcI857CFP and λcI857YFP carrying pBAD18 and uninfected E . coli MG1655 carrying plasmid pRSET-mCherry . 20μg/ml Ampicillin was added to the medium to prevent spontaneous plasmid loss . Horizontal transmission events were scored by double coloration ( mCherry+CFP and mCherry+YFP ) . Horizontal and vertical fitness were calculated as mentioned above by only considering mCherry+ or mCherry- cells . The relative fitness of λcI857 was calculated as W = ( ft / ( 1- ft ) ) / ( f0 / ( 1- f0 ) ) where ft measures the frequency at time t . In general the frequency was calculated over the whole host population . In Fig 6 we distinguish horizontal fitness , WH , and vertical fitness , WV , by focusing either on the frequency in newly infected bacteria ( bacteria carrying the plasmid pRSET-mCherry ) or on the frequency in previously infected bacteria ( bacteria that do not carry the plasmid pRSET-mCherry ) , respectively . In Fig 4 we present the ratio WH / WV for different mixing treatments . Analyses were carried out using the R statistical package ( version 3 . 0 . 2 ) . The fitness W of λcI857 ( Fig 5 , S1 Data ) was log transformed prior to fitting in a mixed effects model ( lme , nlme package ) . Mixing ( undisturbed , 30 s , 24h and 24h-wet ) and transfer day ( 1 , 2 and 3 ) were fitted as fixed effects while population was nested within the marker effect ( the two reciprocal marker-mutant combinations ) to account for the repeated measurements at different points in time . The ratio WH / WV ( Fig 6 , S1 Data ) was fitted in a mixed effect model with mixing as a fixed factor and marker as a random effect . The prevalence of the infection ( Fig S3 in S1 Text and S1 Data ) was fitted in a mixed effect model with mixing as a fixed factor and marker colour as a random effect . Maximal models , including all higher-order interactions , were simplified by sequentially eliminating non-significant terms and interactions to establish a minimal model . The significance of explanatory variables in mixed effect models was established using a likelihood ratio test which is approximately distributed as a chi-square distribution [40] . A posteriori contrasts were carried out at the final time point of the experiments in Figs 5 , 6 and S3 ( in S1 Text ) by aggregating factor levels together and by testing the fit of the simplified model using a likelihood ratio test [41] . | Why do some viruses use latent infection strategies and in which environments can such prudent host exploitation strategies evolve ? Theory of mathematical epidemiology and social evolution predicts that prudent pathogens can evolve when they cluster in space and share the mutual benefits of prudent exploitation of the host population . Here we extend the theory to study the effect of spatial structure on the evolution of pathogens that can transmit both horizontally and vertically . We explore these effects experimentally by competing the latent bacteriophage λ and the virulent mutant λcI857 in spatially structured epidemics and gradually erode spatial structure . We show that the latent bacteriophage λ only wins in a spatially structured epidemic . Yet , a breakdown of the epidemic structure by long-range transmission reduces the benefit of latency by 500 fold . This demonstrates that long-rage transmission , as for example by air travel , might select for much more virulent types of previously latent viruses . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [] | 2015 | Spatial Structure, Transmission Modes and the Evolution of Viral Exploitation Strategies |
Clonorchis sinensis causes chronic cumulative infections in the human hepatobiliary tract and is intimately associated with cholangiocarcinoma . Approximately 35 million people are infected and 600 million people are at risk of infections worldwide . C . sinensis excretory-secretory products ( ESP ) constitute the first-line effector system affecting the host-parasite interrelationship by interacting with bile fluids and ductal epithelium . However , the secretory behavior of C . sinensis in an environment close to natural host conditions is unclear . C . sinensis differs from Fasciola hepatica in migration to , and maturation in , the hepatic bile duct , implying that protein profile of the ESP of these two trematodes might be different from each other . We conducted systemic approaches to analyze the C . sinensis ESP proteome and the biological reactivity of C . sinensis glutathione transferases ( GSTs ) , such as global expression patterns and induction profiles under oxidative stress and host bile . When we observed ex host excretion behavior of C . sinensis in the presence of 10% host bile , the global proteome pattern was not significantly altered , but the amount of secretory proteins was increased by approximately 3 . 5-fold . Bioactive molecules secreted by C . sinensis revealed universal/unique features in relation to its intraluminal hydrophobic residing niche . A total of 38 protein spots identified abundantly included enzymes involved in glucose metabolism ( 11 spots , 28 . 9% ) and diverse-classes of glutathione transferases ( GSTs; 10 spots , 26 . 3% ) . Cathepsin L/F ( four spots , 10 . 5% ) and transporter molecules ( three spots , 7 . 9% ) were also recognized . The universal secretory proteins found in other parasites , such as several enzymes involved in glucose metabolism and oxygen transporters , were commonly detected . C . sinensis secreted less cysteine proteases and fatty acid binding proteins compared to other tissue-invading or intravascular trematodes . Interestingly , secretion of a 28 kDa σ-class GST ( Cs28σGST3 ) was significantly affected by the host bile , involving reduced secretion of the 28 kDa species and augmented secretion of Cs28σGST3-related high-molecular-weight 85 kDa protein . Oxidative stressors induced upregulated secretion of 28 kDa Cs28σGST3 , but not an 85 kDa species . A secretory 26 kDa μ-class GST ( Cs26μGST2 ) was increased upon treatment with oxidative stressors and bile juice , while another 28 kDa σ-class GST ( Cs28σGST1 ) showed negligible responses . Our results represent the first analysis of the genuine nature of the C . sinensis ESP proteome in the presence of host bile mimicking the natural host environments . The behavioral patterns of migration and maturation of C . sinensis in the bile ducts might contribute to the secretion of copious amounts of diverse GSTs , but a smaller quantity and fewer kinds of cysteine proteases . The Cs28σGST1 and its paralog ( s ) detoxify endogenous oxidative molecules , while Cs28σGST3 and Cs26μGST2 conjugate xenobiotics/hydrophobic substances in the extracellular environments , which imply that diverse C . sinensis GSTs might have evolved for each of the multiple specialized functions .
Clonorchis sinensis is a trematode parasite that resides in the hepatobiliary tract of mammals including humans . Its enzootic infection is highly prevalent in several Asian countries including China , Vietnam , Japan and Korea . Approximately 35 million people are infected and 600 million people are at risk of infections worldwide [1] , [2] . Humans are infected with C . sinensis by eating raw or undercooked cyprinoid fish harboring the infective metacercariae . When humans are lightly infected with C . sinensis , most are asymptomatic . However , chronic cumulative infections cause several symptoms associated with biliary ductal systems such as mechanical obstruction and stone formation . Histopathological alterations include ductal dilatation and periductal fibrosis combined with an adenomatous metaplasia . The most serious complication of clonorchiasis is associated with cholangiocarcinoma ( CCA ) [1] , [3] . CCA results from malignant transformation of cholangiocytes . CCA is the second most common type of primary liver cancer [4] , [5] . The etiology of CCA remains largely undetermined , but evidence indicates that the primary sclerosing cholangitis , parasitic infections and hepatitis are predisposing factors [6] . Epidemiological studies have convincingly demonstrated relationships between clonorchiasis and CCA; Clonorchis-infected individuals might die earlier than non-infected individuals due to the high incidence of CCA in areas where clonorchiasis is prevalent [7] , [8] . C . sinensis has been classified as a Group 1 biological carcinogenic agent [9] . C . sinensis thrives for over 10 years in humans . To ensure its long survival within the bile ducts , where hydrophobic substances , immune effector molecules and several kinds of glycosylated host enzymes are profuse [10] , the fluke continuously releases bioactive molecules to cope with the cytopathic environments . The proteinaceous and non-proteinaceous components secreted by the parasite , referred as excretory-secretory products ( ESP ) , are intimately involved in biological processes including the induction of immune compromise and evasion , parasite feeding and detoxification/neutralization of host-derived cytotoxic molecules . Effects of helminth ESP might be reciprocal . ESP might promote proliferation of epithelial cells through its stimulant effects , which accompanies changes in gene expression patterns [11]–[13] . Conversely , ESP might contribute to the creation of an anti-inflammatory micromilieu , where hostile host effector cells prevail . Antioxidant enzymes , molecules involved in local immune regulation and cysteine proteases likely play critical roles during these interactions [14]–[16] . Glutathione transferases ( GSTs ) are a group of highly versatile and multifunctional enzymes that principally function in conjugation of glutathione to electrophilic donors , thereby participating in the detoxification processes , especially in helminth parasites . Several different types of acidic/neutral cytosolic- , μ- , σ- and ω-class GSTs have been characterized in helminth parasites [17] . Most helminth GSTs are expressed in the parasite's parenchyme and subteguments , which suggests either that they interact with physiologically/pathologically active molecules during the phase II detoxification process , or that these enzymes are secreted and directly contact the host's defensive system [18] . Crystal structure and molecular modeling studies have revealed critical differences in their substrate and/or inhibitor specificity according to xenobiotic substrate-binding sites [19] , [20] , which implies that these isozymes might mediate different cellular homeostatic processes . A total of four GST isozymes have been recognized in C . sinensis . A 28 kDa σ-type CsGST ( Cs28GST; AF051318 ) is expressed in the subtegument and underlying mesenchymal tissues [21] . Two cDNAs that encoding cytosolic 28 kDa GSTs ( CsGST , DQ179264 and CsGST1 , DQ342327 ) , whose substrate specificity and inhibitor profiles of the recombinant enzymes were partly analyzed , have been isolated , but their exact identities remain elusive [22] , [23] . A 26 kDa GST ( cs26GSTM; L47992 ) , having a folding topology similar to the μ-class GSTs , can conjugate GSH to the reactive carbonyl of peroxidized lipids [24] . ( Protein names of each GST are adapted from original authors ) . However , no systemic approaches have yet been undertaken to analyze the biological reactivity of CsGSTs , such as global expression patterns and induction profiles under oxidative stress and host bile . Moreover , the secretory behavior of C . sinensis in an environment close to natural host conditions has remained with uncertainty . In the present study , we comparatively analyzed the ex host C . sinensis ESP proteome obtained in the presence and absence of the rabbit bile . ESP obtained in the presence of the host bile might represent the authentic nature of the C . sinensis ESP , because C . sinensis resides in the bile-filled hepatobiliary ducts of a definitive host . C . sinensis differs from Fasciola hepatica in migration to , and maturation in , the hepatic bile duct , implying that the proteome array of C . sinensis ESP might be different from that of F . hepatica . We observed that incubation medium supplemented with host bile substantially augmented the parasite's secretion , but retained its compositional profile . Interestingly , release of σ- and μ-classes of GSTs was differentially regulated in response to the host bile and oxidative stressors . We provide the basis for identification of genuine C . sinensis excretory-secretory proteome , which might be critically involved in the pathobiological alterations , thus significantly deepening our understanding of this clinically important human pathogen .
All animals were housed in accordance with guidelines from the Association for the Assessment and Accreditation of Laboratory Animal Care ( AAALAC ) . All protocols for animal infections were approved by the Institutional Review Board and conducted in the Laboratory Animal Research Center of Sungkyunkwan University ( protocol 2008-8-18 ) . C . sinensis metacercariae were collected from naturally infected freshwater fish , Peudorasbora parva , in an endemic area in Jinju , Gyeongsangnam-do , Korea . Three New Zealand White rabbits were orally infected with 500 metacercariae/rabbit using a gavage needle . The adult worms were recovered from rabbit bile ducts at 8-week post-infection . The worms were washed more than 10 times with physiological saline at 4°C . Bile fluid was aseptically drawn from gall bladders of three age-matched rabbits and centrifuged at 20000 g for 30 min at 4°C to remove host-derived cell debris . Viable intact worms were collected under a dissecting microscope . The worms were placed in serum-free RPMI-1640 medium ( pH 7 . 2; Life Technology ) at 37°C for 1 h to ensure emptying of the guts . The flukes were incubated in serum-free RPMI-1640 medium in the absence or in the presence of rabbit bile ( 1% , 5% , 10% or 20% ) or hydrogen peroxide ( 0 . 5 mM ) at 37°C for 1 h ( 20 worms/group/3 ml medium ) . Incubations of dead but intact worms , which were prepared by treating the worms for 10 min at 55°C in RPMI-1640 medium with/without 10% host bile , were included as negative controls . After removing the worms , the conditioned medium was centrifuged at 3000 g for 5 min followed by 20000 g for 30 min at 4°C . Resulting supernatants were used as ESP . After incubation , the worms were washed three times with ice-chilled PBS ( 100 mM , pH 7 . 2 ) and were immediately used in the extraction of proteins and RNAs . All procedures were repeated in triplicate . ESP were also prepared in a large scale ( 10 ml ) with 100 worms , to obtain sufficient samples for the identification of secreted proteins as well as to minimize variation caused by different physiological conditions of individual worms . Protein profiles in an equal volume and/or an equal amount of protein contents of the experimental ESP samples were examined by 12% SDS-PAGE under reducing conditions and/or by 2-dimensional electrophoresis ( 2-DE ) as described below . ESP were precipitated with ice-chilled 10% tricholoacetic acid/acetone and resuspended with rehydration buffer containing 6 M urea , 2 M thiourea , 2% CHAPS , 0 . 4% DTT , 0 . 5% IPG buffer ( pH 3–10; GE Healthcare ) and 0 . 002% bromophenol blue ( BPB ) . The proteins were separated along with the respective isoelectic points on IPG strips with a cup-loading instrument ( Amersham Biosciences ) for a total of 35 kVh , followed by 12% SDS-PAGE ( 160×160×1 mm ) . Protein spots were stained with colloidal Coomassie Brilliant Blue G-250 ( CBB ) . The spots of interests were excised from the 2-DE gels and were processed with in-gel tryptic digestion . The tryptic peptide mixtures were analyzed with an AutoflexII matrix-assisted laser desorption/ionization-time of flight mass spectrometry ( MALDI-TOF MS ) system ( Bruker Daltonics , GmbH ) . The peptide mass fingerprints ( PMFs ) and tandem MS obtained from each spot were searched against the NCBI protein sequence database ( http://www . ncbi . nlm . nih . gov ) and locally mounted C . sinensis EST and draft genome database [25] , [26] using Biotools software ( Bruker Daltonics ) . The PMF tolerance was ±100 ppm and tandem MS tolerance was ±0 . 8 Da . Single missed cleavage site , cysteine carbamidomethylations and methionine oxidation were considered . Protein identifications were based on Mascot scores that were above the significant threshold , which was set at p<0 . 05 . Protein probability was assigned by the Protein Prophet algorithm [27] . For the MALDI-TOF/TOF MS analysis , all of the spectra were obtained using a long-lifetime N2 laser , operating in nitrogen flow-through mode ( 20 Hz frequency ) , yielding laser irradiation at λ = 337 nm with a maximal laser power of approximately 110 µJ . Typically 500–800 shots of MS acquisition were summed for the acquisition of a laser-induced dissociation spectrum at elevated laser power . Bruker Biotools software was used to interpret and annotate the tandem MS . Database searches were conducted with the Matrix Science Mascot search engine program . C . sinensis genomic DNA was isolated using a Wizard Genomic DNA Purification Kit ( Promega ) . The chromosomal segments of Cs28σGSTs ( AF051318 and DQ342327 ) were amplified from the genomic DNA employing long-range PCR . The primers were designed from the nucleotide sequences within both ends of each cDNA ( CsGSTs-g-F and CsGSTs-g-R; Table S1 ) . The amplification reactions were done with the LA Taq system under a thermal cycling profile recommended by the manufacturer ( Takara ) and analyzed on an agarose gel . The amplicons cloned into pGEM-T Easy vectors ( Promega ) were sequenced and compared with their corresponding cDNA sequences to determine the positions and lengths of intervening introns . The general rules for splice site consensus sequences were considered to verify the accuracy of exon-intron boundary regions ( gt-ag ) . O . viverrini DNA was isolated from the lyophilized worms , which were a generous gift from Dr . B Insisiengmay ( Department of Hygiene and Prevention , Ministry of Public Health , Lao PDR ) . The genomic structure of O . viverrini ortholog ( AY057838 ) was similarly determined using the genomic DNA and a primer pair ( Ov28GST-g-F/R; Table S1 ) . The open reading frame ( ORF ) regions of Cs26μGST and Cs28σGST genes amplified by PCR from an adult C . sinensis cDNA library with gene-specific primers ( Table S1 ) were cloned into the pET-28a expression vector and transformed into competent Escherichia coli BL21 ( DE3 ) . The expression of recombinant proteins was induced by adding 0 . 5 mM isopropyl-1-thio-β-D-galactopyranoside ( IPTG ) into the bacterial cell cultures . The recombinant proteins were purified by nickel-nitrilotriacetic acid ( Ni-NTA ) affinity chromatography ( Qiagen ) and monitored by 12% reducing SDS-PAGE . Polyclonal antibodies were generated by immunizing BALB/c mice with respective recombinant GST proteins mixed with Freund's adjuvants ( Sigma-Aldrich ) according to the standard protocol . Final booster was done by intravenous injection of 10 µg proteins . The blood was collected by heart puncture , centrifuged for 10 min at 3000 g and the respective antisera were stored at −80°C until use . Fresh intact worms were pre-incubated in 50 ml serum-free RPMI-1640 medium for 1 h at 37°C . The worms ( 20 worms/group/3 ml medium ) were transferred into the fresh medium containing methyl viologen dichloride hydrate ( Paraquat; 25 and 100 mM ) , 5-hydroxy-1 , 4-naphthoquinone ( Juglone; 25 and 100 µM ) or hydrogen peroxide ( H2O2; 0 . 5 and 2 mM ) , followed by 1 h incubation at 37°C . The soluble proteins from the respective conditioned worms were separately prepared by homogenizing the worms in PBS ( 100 mM , pH 7 . 2 ) followed by centrifugation at 20000 g for 30 min at 4°C . The conditioned medium and proteins extracted from worms incubated without oxidative treatment were taken as controls . The induction profiles of CsGSTs were examined through Western blots probed with each of the CsGST-specific mouse antisera . An anti-cathepsin F antibody available in our laboratory was used as a quantity/quality control , of which constant expression was verified by reverse transcription ( RT ) -PCR , in association with the respective stimuli ( unpublished observation ) . Intact viable adults were also incubated with rabbit bile ( 1–20% ) for 1 h at 37°C and the induction profiles of CsGSTs were assessed with the worm extracts and ESP . Three independent experiments were done with freshly prepared worms . Proteins separated by SDS-PAGE/2-DE were transferred to nitrocellulose membranes ( Schleicher & Schuell , BioScience ) . The membranes were blocked in Tris-buffered saline ( 100 mM , pH 7 . 6 ) supplemented with 1% Tween 20 ( TBST ) and 5% skim milk , after which membranes were incubated with the respective anti-CsGST antibodies ( 1∶1000 dilution of each ) and subsequently with horseradish peroxidase-conjugated goat anti-mouse IgG antibody ( 1∶4000 dilution ) . The signals were detected using an ECL detection system ( Amersham Biosciences ) . For quantitative analysis , all images were exposed for 1 min . Total RNAs were extracted from adult C . sinensis with the TRIzol reagent ( Invitrogen ) . First-strand cDNA was synthesized by reverse transcribing 3 µg of total RNAs in a final reaction volume of 100 µl using a RNA PCR Kit ( AMV , ver2 . 1; Takara ) and an oligo ( dT ) 15 primer . The cDNAs were used in the following PCR procedure to amplify each of the antioxidant genes in 96-well plates . The PCR mixtures contained 5 µl of diluted cDNAs ( 1∶5 ) , 5 µl of 2× SYBR Green PCR Master Mix ( Applied Biosystems ) and 200 nM of each gene-specific primer ( Table S1 ) . Real-time qRT-PCRs was performed in a thermocycling profile ( 95°C for 5 min followed by 40 cycles of 15 s at 95°C and 1 min at 60°C ) employing ABI Prism 7000 Sequence Detection System and software ( PE Applied Biosystems ) . Control reactions were done with RNAs that had not been reverse transcribed to verify absence of genomic DNAs in the RNA samples . All of the reactions were independently conducted in triplicate . The specificity of the amplicons was verified by melting curve analysis at a temperature ranging from 60–95°C , after observing the PCR products by agarose gel electrophoresis . The data were normalized against the values obtained with the β-actin gene ( EU109284 ) and the fold-induction in the experimental groups was calculated by comparison of those of the non-stimulated controls .
We incubated fresh intact adult worms in RPMI-1640 at 37°C for 1 h in the presence or absence of host bile . The conditioned media were examined by 12% SDS-PAGE to observe the effects of host bile on C . sinensis excretory-secretory behavior . C . sinensis released numerous proteins that were mainly 18–30 kDa and 50–90 kDa . The banding profiles were highly reproducible in three different batches according to increasing doses of host bile ( a representative data set is shown in Figure S1A ) . However , the relative intensity of each band and the total amounts of the proteins were substantially increased in the bile-induced ESP in a dose-dependent manner , in which approximately 3 . 5- and 5-fold increase were observed in the 10%- and 20%-bile ESP , respectively ( Figure S1B ) . Addition of 0 . 5 mM H2O2 into the incubation medium also augmented the protein secretion , especially of the 20–30 kDa species . Time lapse analysis indicated that incubation of C . sinensis up to 8 h in RPMI-1640 did not induce considerable alteration in the protein profile , but the total amount of ESP and the relative intensity of the respective bands were substantially increased in proportion to the incubation time ( data not shown ) , as previously observed with Fasciola hepatica [13] , [28] . We detected more than 100 protein spots on CBB-stained 2-DE gel of normal ESP ( ESP obtained by incubating the worms in the absence of host bile ) , in which 200 µg proteins were loaded ( Figure 1A ) . These spots were evenly distributed across a broad pH range of 5–10 and the relative intensities were similar to one another , except for spots 22 , 27 , 40 and 41 . The profile appeared to be relatively well maintained in the same concentrations of 10%-bile ESP , although the spots in the acidic pH region became complicated by superimposition of bile proteins ( Figures 1D and 1E ) . Total amounts of proteins in normal and bile-induced ESPs showed a certain degree of variation in the three separate preparations ( <20% ) . Nevertheless , overall spotting patterns were comparable among the preparations ( data not shown ) . These collective data suggested that global excretion-secretion pattern of C . sinensis was not significantly affected by the elapsed time either after being removed from the host or after being exposed to the in vitro stresses . Rabbit bile , which comprises the natural host environment of C . sinensis , was cytopathic against the parasite , even when used in concentrations of 1–20% ( v/v ) , probably because the bile juice collected from normal rabbit gall bladder were enriched with secretory IgA and mucin of up to 10-fold concentration [29] . We maintained in vitro incubations for 1 h in the presence of 10% bile juice to minimize the adaptive changes and degeneration of the worms during incubation in ex host conditions . The protein identities were determined by MALDI TOF/TOF MS or MALDI TOF-MS , and following Mascot analyses against the non-redundant proteomic database of GenBank and locally mounted EST and draft genome database [25] , [26] . A total of 193 protein spots ( 138 from 10%-bile ESP and 55 from normal ESP ) together with 15 from 10%-bile control were selected for the identification . Most spots showed a good PMF score , but many spots could not be precisely identified during the searches . Moreover , although a great deal of information exists in genomic and proteomic databases for S . japonicum and F . hepatica , these databases did not match well with protein spots detected in C . sinensis ESP . We were able to identify 38 protein spots from the normal and 10%-bile ESP , of which three protein spots ( 5 , 12 and 15 ) were originated from rabbit and six proteins from 10%-bile control ( spots a–f , Figure 1C ) . The C . sinensis proteins were largely categorized into three major groups: Glycolytic enzymes ( 11 spots , 28 . 9% ) , antioxidant GSTs ( 10 spots , 26 . 3% ) and proteolytic enzymes ( four spots , 10 . 5% ) ( Table 1 . See also Table S2 for detailed description ) . The primary fractions of the proteins contained enzymes involved in the glycolysis such as enolase ( spots 6 and 8–10 ) , glyceraldehyde-3-phosphate dehydrogenase ( spot 17 ) and triosephosphate isomerase ( spots 28 and 29 ) , or in Krebs cycle ( malate dehydrogenase; spot 19 ) . Spot 7 was related with F . hepatica enolase , but the Mascot value of 37 was not significant ( significant value is >43 ) . Mitochondria-type aldehyde dehydrogenases ( spots 3 and 4 ) , which are responsible for the conversion of toxic aldehydes into non-toxic compounds , and succinate-semialdehyde dehydrogenase ( spot 18 ) , which participates in the degradation of γ-aminobutyric acid , were additionally identified . A total of 10 GST family members including the μ- , σ- and ω-classes were detected in the molecular weights ( Mr ) and isoelectric point ( pI ) ranges of 26–28 kDa and 5 . 2–9 . 3 , respectively ( Figures 1D and 1E ) . The PMF of the 28 kDa molecules matched to those of previously identified C . sinensis GSTs; spot 22 ( pI 5 . 2 ) to CsGST1 ( ABC72085 ) , spot 32 ( pI 7 . 5 ) to CsGST ( ABA56496 ) , and spots 33 and 34 ( pIs 8 . 4 and 9 . 0 ) to Cs28GST ( AAD17488 ) ( Table 1 and Table S2 ) . Protein spot 22 ( pI 5 . 2 ) was identical to parenchymal-type CsGST1 ( ABC72085 ) [23] . This protein was not confined intracellularly , but was secreted . Interestingly , its expression was significantly decreased in the presence of 10% host bile . Spot 32 ( pI 7 . 5 ) was identified as another σ-class GST ( CsGST; ABA56496 ) [22] . Spots 33 and 34 ( pI 8 . 4 and 9 . 0 ) were closely related to Cs28GST ( AAD17488 ) [21] . We renamed these σ-type GSTs as Cs28σGST3 ( acidic protein ) , Cs28σGST2 ( neutral protein ) and Cs28σGST1 ( basic proteins ) , respectively , in accordance with their discovery order . The multiple 26 kDa proteins with pI 6 . 5 ( spots 35–37 ) were identical to μ-type cs26GSTM1 ( AAB46369 ) [24] . Spot 23 ( pI 5 . 2 ) displayed a PMF that matched to Echinococcus multilocularis μ-type GST ( CAA59739 ) . We designated spots 35–37 as Cs26μGST2 and spot 23 as Cs26μGST1 . The PMF analysis further identified spots 24 and 25 orthologous to the Schistosoma mansoni ω-type GST ( AF484940 ) [30] , which was newly recognized in C . sinensis ESP proteomes ( Figure 1D and Table 1 ) . The cathepsin L-like cysteine proteases were a common type of proteolytic enzyme released by C . sinensis ( spots 16 , 27 , 30 and 31 ) . Their Mr ( 26–28 kDa ) indicated these proteases were secreted after proteolytic maturation; however , spot 16 , which was detected only in bile ESP , had a large Mr ( >32 kDa ) and exhibited PMF matched to the ‘Inhibitor I29’ ( 42TYSNDDDELRFEIFK56 ) conserved in the prodomain of cathepsin F [31] . Each of two spots homologous to paramyosin , myosin and aldehyde dehydrogenase , as well as one spot each representing ferritin , myoglobin and fatty acid-binding protein were detected . Protein spots of 13 and 20 were associated with S . mansoni myosin II heavy chain ( CAA46548; Mr/pI of 61585/5 . 20 ) and S . japonicum metalloprotease ( AAX27148; Mr/pI of 32683/8 . 43 ) , while they showed only one peptide hit . In addition , a portion of the tryptic peptides obtained from some other protein spots revealed good PMF score or MS spectra , but homologous identities could not be retrieved from the GenBank databases due mainly to their low sequence coverage . To further validate the applicability of our in vitro incubation system , we determined the protein profile of 10%-bile control . Multiple rabbit proteins released into bile were detected , with the identities of poly-immunoglobulin-receptor , albumin , annexin A , adhesion molecules and immunoglobulin J-chain ( Figure 1C and Table 1 ) . These proteins , except for spots e and f , were hardly detected in the bile ESP , probably due to the proteolytic degradation induced by Clonorchis proteases and/or to the superimposition of numerous proteins within the narrow acidic region ( Figure 1B ) . Analysis of primary and tertiary structure of Cs28σGSTs revealed that amino acid residues involved in the GSH binding and enzymatic catalyst were highly conserved in the Clonorchis protein and their homologs ( Figure S2 ) . Cs26GSTs exhibited distinct patterns of amino acid conservation with several μ-type GSTs ( data not shown ) similar to those of schistosome orthologs [32] . Nevertheless , the recombinant forms of Cs26μGST2 , Cs28σGST1 and Cs28σGST3 demonstrated comparable biochemical properties in regard to pH and temperature optima and inhibitor sensitivity ( Figure S3 ) . The close relationship between trematode μ- and σ-class GSTs was further evaluated by a cladistic analysis of 239 GST members retrieved from the protostomian databases of the GenBank , using the neighbor-joining algorithm of MEGA . The phylogenetic tree separated these proteins into distinct clades , consistent with the types classified by the original authors ( Figure S4 ) . The major portion of the protostomian GSTs deposited in the databank seemed to belong to the σ-class , which were allocated well into separate clusters according to the taxonomies of their donor organisms at the class level . The platyhelminth GSTs positioned in the clades proximal ( σ- , α- and π-like proteins ) or identical ( μ-like proteins ) to the μ- and π-class members of ecdysozoans , although some of the branching nodes were not statistically significant . This unique phylogenetic relationship would partly support the overlapping enzymatic properties of several trematode GSTs . The phylogenetic analysis suggested that the genic dosage of σ-type GST has been preserved differently among trematode genomes . Analysis of S . mansoni and S . japonicum genomic databases in the Sanger Institute ( http://www . sanger . ac . uk/ ) and Shanghai Center for Life Science & Biotechnology Information ( http://lifecenter . sgst . cn/ ) indicated the single-copy dosage of schistosome orthologs ( data not shown ) , while the copy number could not be properly estimated in C . sinensis . The evolutionary relationships among the trematode σ-class GST genes were further examined by comparing their genomic structures , which were isolated from respective genomes either by PCR amplification ( C . sinensis and O . viverrini ) or in silico screening ( S . mansoni ) . These genes exhibited an identical structural topology with four exons and three introns . Their respective intervening introns occupied orthologous positions , despite the significant length polymorphism ( Figure 2A ) . BLAST searches detected a SjR2-like non-long terminal repeat ( LTR ) retrotransposon [33] sporadically integrated in the first intron of Cs28σGST3 , which contributed to the length polymorphism of Cs28σGST3 . Southern blot analysis of C . sinensis genomic DNAs with multiple probes specific to various regions of Cs28σGST3 demonstrated that the other introns were also evolved via insertional events of repetitive genetic elements of as-yet-unidentified repetitive elements ( Figure 2B ) . The chromosomal segment of Sm26μGST was composed of seven exons and six introns , but none was orthologous to those of the σ-type GST genes ( data not shown ) . These collective data suggest that either C . sinensis σ-class GST lineage has undergone duplication events to expand this paralogous gene pool in C . sinensis , unlike schistosome genomes . A previous study demonstrated that Cs28GSTs might play a major role in antioxidant activity compared to Cs26GSTs because molar ratio of these two proteins approximated 14∶1 in adult C . sinensis [21] . The protein array revealed that σ-class GSTs constituted the major GST fractions secreted by C . sinensis in the absence of bile juice ( Figure 1D ) . The conditions were parallel with global GST expression profile of C . sinensis parenchyme , which was purified using a GSH-affinity column and following 2-DE analysis ( Figure 3B ) . Conversely , 10%-bile ESP showed a critical difference , in which the release of Cs28σGST3 ( spot 22 ) was significantly decreased , while those of Cs26μGST2 ( spots 35–37 ) were increased ( Figure 1E ) . These results suggested strongly that the functional expression and secretion of GSTs might be different in response to exogenous stimuli including bile . To delineate the expression pattern of these GSTs , we cloned Cs28σGST1 and 3 ( AAD17488 and ABC72085 ) and Cs26μGST2 ( AAB46369 ) , which exhibited significantly upregulated or downregulated secretion in the presence or absence of host bile . We expressed these GSTs in E . coli cells and generated monospecific antibodies that did not cross react with one another . We examined the expression pattern of these GSTs in the normal ESP ( 20 µg ) and the worm extracts ( 50 µg ) employing these antibodies . As shown in Figure 3A , when worm extracts were probed with anti-recombinant Cs28σGST3 antibodies , protein spots of ca . 56 and 85 kDa were detected in addition to a prominent 28 kDa spot ( The reactions at 41 and 47 kDa were not consistent in other experiments ) . The 28 kDa protein was abundantly expressed in the parenchyme and , in part , released into the ESP ( Figure 3A , panels a and d ) . Cs28σGSTs with pIs 9 . 0 and 9 . 3 were the major fractions in the parenchyme , while those with pIs 8 . 0 and 8 . 3 constituted main ESP fractions . Considering the identification profiles ( Table 1 ) and the tightly conserved regions within their primary structures ( Figure S2A ) , spots seen above pH 9 . 0 would be Cs28σGST1-related proteins and those with pH 8 . 0 were likely to be homologous to Cs28σGST2 ( Figure 3A , panels b and e ) . A small amount of isoforms generated via posttranslational modification , or unidentified paralogs might be secreted , as seen by their different reaction intensities at different pIs on Western blotting . Cs26μGST2 fractions were located at neutral pH ( 6 . 5–7 . 0 ) and most of these proteins appeared to be secreted ( Figure 3A , panels c and f ) . Due to highly complicated pattern including superimposition of proteins of the bile ESP , we were unable to identify a GST-related high-molecular-weight-protein at 85 kDa ( Figure 1B ) , which were clearly observed in immunoblotting ( Figures 3A , panels a and d , denoted by asterisks ) . The global CsGST fractions were purified through a GSH-affinity column from the worm extracts . The 26 and 28 kDa proteins separated into multiple spots according to their respective pI values ( Figure 3B ) , which showed similar patterns to those of excretory GSTs in the absence of bile . These spots were indeed identified as the same molecules identified in the ESP ( spot numbers are same as in Figure 1 ) . We detected an 85 kDa protein with an acidic pI ( Figure 3B , asterisk ) . The protein was identified as Cs28σGST3 by MALDI-TOF/TOF MS analysis ( Figure 3C ) . The protein also demonstrated a high PMF score with the O . viverrini 28 kDa GST ( Ov28GST; AL2371 ) . Spot ( s ) corresponding to Cs28σGST2 could not be detected in the fractionated GSTs , probably owing to the relatively low expression level of the protein ( Figure 3A , panel b ) . The protein spots corresponding to the secretory ω-class GSTs ( spots 24 and 25; Figures 1D and 1E ) were not purified by GSH-affinity chromatography ( Figure 3B ) , which was relevant to the minimal affinity of the class members toward agarose-coupled GSH [30] . The protein array of C . sinensis ESP indicated that diverse GSTs are released into the surrounding environments , which was strongly suggestive of their primary defensive roles in protecting the worms from exogenous toxic substances . The secretory CsGSTs were substantially increased in bile-induced ESP . However , the rabbit bile seemed to suppress the release of Cs28σGST3 , which obviously contrasted to its upregulatory effects of secretion of other GST molecules ( Figures 1D and 1E ) . We assessed the induction profile of the CsGSTs upon stimulation with oxidative chemicals . Oxidative stress conditions created by H2O2 and juglone significantly increased the release of Cs28σGST3 and Cs26μGST2 in a dose-dependent manner . Paraquat , a superoxide free radical generator , augmented secretion of Cs28σGST3 and Cs26μGST2 at low concentrations ( 25 mM ) , but inhibited their secretions at high concentrations ( 100 mM ) . However , secretory Cs28σGST1 was not induced in response to these oxidative stressors ( Figure 4A , upper panels ) . The expression pattern of these three GST molecules was relatively constant in intracellular compartments ( Figure 4A , lower panels ) . The decreased expression of Cs28σGST3 and Cs26μGST2 in the worms incubated with 100 mM paraquat might result from the lethal effects of the chemical at the high concentrations ( approximately 60% of the worms survived after the treatment ) . We subsequently examined the excretion of GSTs upon treatment with increasing doses of H2O2 . The secretion of Cs28σGST3 and Cs26μGST2 were increased significantly in a dose-dependent fashion , while that of Cs28σGST1 was hardly observed ( Figure 4B ) . Secretory cathepsin F , which was used as a negative control , showed no expressional changes upon the oxidative stimuli . These collective data demonstrated that different GST species exerted their differential roles in response to different exogenous stimulus . To scrutinize the expression and secretion pattern of CsGSTs upon bile stimulation , worm extracts obtained from cultures with rabbit bile were analyzed by Western blotting . As bile concentrations increased , the secretion of 85 kDa Cs28σGST3 species dose-dependently increased , whereas that of 28 kDa Cs28σGST3 decreased ( Figure 5A , left panel ) . In addition , the 28 kDa Cs28σGST3 in the parenchyme appeared to be proportionally decreased along with increased secretion of 85 kDa Cs28σGST3 species . The expression of Cs28σGST1 was slightly augmented in the parenchyme , while secretion of Cs28σGST1 was negligible ( Figure 5A , middle panel ) . Secretion of Cs26μGST2 was also increased to some extent , while its expression was not significantly modified in the intracellular compartment ( Figure 5A , right panel ) . We then assessed whether the substantial increase of the secreted 85 kDa Cs28σGST3 was due to an abiotic cross-linking of the Cs28σGST3 monomer under the high concentrations of bile . We coincubated C . sinensis normal ESP with 10% and 20% rabbit bile for 1 h at 37°C , but found no significant polymerization effect ( Figure 5B ) . We further analyzed whether this phenomenon was associated with the altered expression of these transcripts by real-time qRT-PCR . As shown in Figure 5C , incubation with the viable worms either in 0 . 5 mM H2O2 or in 1–20% rabbit bile did not significantly revise the expression of these transcripts , although those of Cs28σGST1 and Cs28σGST3 marginally increased and that of the Cs26μGST2 was slightly decreased as bile concentration increased . These results indicated the relative stability of the expression levels of these GSTs , but that C . sinensis might adjust its secretion behavior to cope with stressful conditions and ensure its survival in the hosts .
In the present study , we incubated fresh viable adult C . sinensis in the presence or absence of host bile in a relatively short period of 1 h to minimize possible alteration of secretory behavior through physiological adaptation during longer term incubation , as well as to mimic the natural host microenvironment . The overall protein expression pattern of C . sinensis was not significantly affected by the host bile , in terms of the spectrum of secretory molecules , but the total amount of secreted protein was substantially increased in the presence of 10% bile up to approximately 3 . 5-fold . This result suggests that bile components might continuously arouse C . sinensis and thus , promote secretion of ESP in the natural host environments . The ESP proteome primarily constituted with several enzymes , whose functions might be related mainly with energy metabolism and nutrition , detoxification , immune evasion and worm migration . Previous studies with C . sinensis and other trematode parasites such as O . viverrini , F . hepatica , S . mansoni , S . japonicum and Paragonimus westermani also demonstrated these proteins are plentiful in the ESP . These collective observations suggest that the adaptive biological processes and associated molecular networks might be tightly conserved in the parasitic trematodes to allow host-parasite interplay [13]–[16] , [28] . Proteome analysis of the 10%-bile ESP demonstrated that some of the major GST proteins such as μ- , σ- and ω-classes were released into the surrounding environment . Secretions of these antioxidant proteins were largely inducible by stimulating worms with host bile or oxidants . However , as is shown in Figures 1D and 1E , secretion of 28 kDa Cs28σGST3 species appeared to be substantially decreased ( up to 95% ) in the bile stimulated ESP , while that of immunologically-related high-molecular weight proteins were profoundly augmented as bile concentrations increased . This is a unique finding observed for the first time with C . sinensis ESP , but not with other trematode such as F . hepatica and O . viverrini , which also reside in the mammalian bile duct [13] , [16] , [34] . We purified the global CsGST fractions using GSH-matrix and observed that the 85 kDa protein shared peptide fragments with the 28 kDa Cs28σGST3 , thus representing a genuine Cs28σGST3 . The high-molecular-weight-protein did not originate from a simple cross-linking process occurring abiotically between the 28 kDa protein and bile components ( Figure 5B ) . Screening of the C . sinensis cDNA library with primers designed from nucleotide sequences of the conserved peptide fragments could not detect any genes with a larger size . Southern blotting of C . sinensis genomic DNA with the Cs28sGST3-specifc probes demonstrated that the parasite genome contained only a single copy gene homologous to Cs28σGST3 cDNA ( Figure 2B ) . Comprehensive analyses on the developmental changes of C . sinensis transcriptome also could not detect putative gene ( s ) encoding the 85 kDa protein homologous to the 28 kDa Cs28σGST3 [35] , [36] . These collective data indicate that the 85 kDa protein represents the multimeric form of the 28 kDa Cs28σGST3 generated by an as-yet-unknown biochemical linkage . A recent study showed that oxidative stress modifies the multimerization of mouse homer scaffolding proteins through disulfide cross-linking [37] . A portion of many of the proteins form oligomeric complexes via inter-molecular disulfide bonds , which can easily be broken by treating the proteins with reducing agents such as DTT and β-mercaptoethanol . However , 85 kDa Cs28σGST3 appeared to be highly tolerant against these reducing agents . Mammalian bile contains enormous hydrophobic substances , such as cholesterol , phospholipids , fatty acids and triglycerides , in addition to proteins and hormones [10] , [29] , most of which influence C . sinensis survival . Therefore , it is conceivable that the parasite might sequester these toxic compounds by conjugating them on the secretory Cs28σGST3 . Further studies are warranted to elucidate the physicochemical cross-linking among 28 kDa molecules or the plausible conjugation mechanism between Cs28σGST3 and bile component ( s ) . S . japonicum ESP abundantly contains a series of fatty acid binding proteins ( SjFABPs ) and heat shock proteins ( HSP90α and HSP70 ) [15] while those of F . hepatica and P . westermani principally possessed diverse isoforms of cysteine proteases , among which cathepsin L/F are the major constituents . These molecules exert their main effects on lipid trafficking and assimilation , parasite feeding , migration and immune evasion in the hosts [13] , [14] , [28] , [38] . Protein array of C . sinensis ESP revealed critical differences compared to those of S . japonicum , F . hepatica and P . westermani , which included secretion of diverse GSTs and relatively low-level secretion of proteases , HSPs and FABPs ( Table 1 ) . However , secretion of several GST isozymes was highly similar to F . hepatica . The different secretory behavior observed in these parasites might reflect the migration manners and favorable localities in their hosts . C . sinensis metacercaria excysts in the duodenum and directly migrates into the bile ducts through the ampulla of Vater within a few days of infection [39] and thrives therein , with its entire lifespan being over 10 years . C . sinensis does not penetrate any host tissues/organs and are hardly exposed to host immune system , which implies that the parasite might secrete effector molecules to attenuate cytopathic bile components . Therefore , the biological significance of the GSTs of C . sinensis and F . hepatica , whose functions are mainly associated with protection from immune cell-derived oxidizing molecules , as well as conjugation , sequestration and detoxification of hydrophobic substance , would be substantially critical to ensure their survival . Secretion of μ- and σ-class GSTs was indeed significantly and specifically accelerated in juvenile F . hepatica that migrated through the liver , compared with the pre- and post-liver stage [40] . On the other hand , F . hepatica and P . westermani travel to their main habitat through the penetration of several internal organs/tissues of the host , during which they continuously interact with host immune effector system [13] , [38] , [40] . The unique migration route of C . sinensis might contribute to secretion of a small quantity and fewer types of cysteine proteases , whose main function is related to tissue hydrolysis and immune evasion . ESP proteome of O . viverrini , which has similar migration behavior and residing niche in the definitive hosts , also revealed significant proportions of antioxidant GSTs but very low amounts of cysteine proteases [16] . C . sinensis secretes copious amounts of myoglobin , which functions as an oxygen reservoir and an intracellular oxygen carrier . Helminth parasites perform anaerobic respiration in the host environments [41] . Probably , C . sinensis expresses a large amount of myoglobin to guarantee maximal uptake of oxygen in the extremely anaerobic environmental niche . Alternatively , the protein may carry out an essential role in scavenging of nitrosative stress molecules generated by host inflammatory cells infiltrating the ductal epithelium [42] . An excretory-secretory myoglobin of Trichostrongylus colubriformis was shown to induce protective immunity against challenging infections following a single intraperitoneal injection [43] . The biological reactivity of C . sinensis myoglobin needs further researches . Long-standing inflammations in the bile ducts , which are associated with helminthic infections such as C . sinensis and O . viverrini , induced hyperplasia and malignant transformation of the cholangiocytes [5] , [9] , [44] , [45] . The inflammatory and oxidative conditions provoked by the parasite infections might trigger genomic instability in hepatobiliary ductal cells , which can lead to the neoplastic transformation via destructive damages on the chromosomal DNAs and proteins [16] , [46] . Cytokines , free radicals and growth factors released by infiltrated host immune cells are also involved in generation of the inflammatory and tumorous conditions [47] , [48] . Finally , restorative hyperplasia in tissues damaged by both direct contact of the parasite to the bile duct epithelium and irritation by the bioactive ESP molecules further promotes the propagation of pre-malignant cells , which are more vulnerable to a carcinogen [45] . Only small proportions of the helminthic infections such as O . viverrini and C . sinensis are related with CCA tumorigenesis . Liver fluke infections remain a major public health problem , especially in high endemic areas , given the large number of people who are exposed to and/or infected [2] , [44] , [49] and the strongest link between CAA compared to other bacterial/viral infections [5] , [8] , [9] . More attention should be paid on the prevention and control of this carcinogenic parasite . In order to investigate secretory proteins concerning their physiological functions , a full range of biochemical , immunological and proteomic analyses should be conducted employing ESP collected in the genuine host environments . However , the natural ESP is not readily collectable from most of parasites , particularly due to the high turnover rate of secreted proteins or their extremely low concentrations [13] , [15] . We incubated C . sinensis in RPMI-1640 medium supplemented with 10% bile for a relatively short period ( 1 h ) to minimize the effects of tegumental turnover and worm degradation [50] . However , we detected multiple proteins , which are believed to be targeted into cytosol such as cytoskeletal proteins ( paramyosin and myosin heavy chain ) and glycolytic enzymes ( enolase , glyceraldehyde-3-phosphate dehydrogenase and triosephosphate isomerase ) . We also detected prodomain of the cathepsin F . Furthermore , a large proportion of proteins identified in the C . sinensis ESP did not contain the classical signal peptides ( Table 1 ) . Similar situations were also observed with O . viverrini [16] , F . hepatica [13] , [28] and S . japonicum [15] . Teguments of platyhelminths including parasitic trematodes have crucial importance for the nutrient uptakes and immune evasion . The secretory granules scattered within the tegument are actively involved in the secretory trail [51] . The trematode proteins might be secreted through the signal peptide-independent secretion mechanism [52] or other unidentified pathway , which is intimately related to the secretory granules . Alternatively , some kinds of proteins , if not all , would be diffused/leaked into the incubation medium through damaged tegument membrane . Quality versus quantity should be carefully taken into consideration when evaluating parasites' ESP . Recent advances in the analyses of the secretory proteome of helminth parasites combined with bioinformatics has immensely widened the understanding of human pathogens by allowing us to identify prodigious numbers of protein molecules , which are decisively involved in the pathobiological alterations [13]–[15] , [28] . Our present study has demonstrated new aspects of host-parasite interactions . The secretory proteome of C . sinensis displayed common and unique features , which might be related to its habitat in the definitive host . The universal secretory proteins found in other parasites such as several enzymes involved in glucose metabolism and oxygen transporters were commonly identified in C . sinensis ESP , whereas cysteine proteases were not abundantly detected . Instead , C . sinensis secreted diverse GSTs , which might be importantly involved in the immunoprotection and detoxification of hydrophobic substances within bile ducts . Our data suggest strongly that different GSTs might have differentially evolved with the specialized functions . Cs28σGST1 and its much closer paralog ( s ) might play major roles as phase II antioxidant enzymes in intracellular compartments , while Cs28σGST3 and Cs26μGST2 may protect worms from exogenously derived toxic substances in extracellular phase , thus shaping the first-line defensive system . | Clonorchis sinensis is a trematode parasite that infects the hepatobiliary ducts of the mammals including humans . Approximately 35 million people are infected and 600 million people are at risk of infections . Epidemiological studies have convincingly demonstrated relationships between clonorchiasis and cholangiocarcinoma . C . sinensis is a Group 1 biological carcinogen . C . sinensis excretory-secretory products ( ESP ) constitute the first-line effector system affecting the host-parasite interrelationship . We observed global expression pattern of C . sinensis ESP in the presence of host bile/oxidative stresses , which mimics natural host environments . The secretory proteome displayed common and unique features , which might be related to its habitat in the definitive host . The universal proteins were found to be several enzymes involved in glucose metabolism , glutathione transferases ( GSTs ) and oxygen transporters . C . sinensis differentially regulated the secretion of diverse GSTs in response to bile and oxidative stressors , which suggested that these enzymes are importantly involved in the protection from immune cell-derived oxidizing molecules and detoxification of hydrophobic substances . C . sinensis secreted less cysteine proteases , which play roles in tissue migration and immune evasion , compared to other tissue-invasive or intravascular trematodes . Our data suggest strongly that different GSTs might have differentially evolved with their specialized functions to cope with stressful conditions and ensure parasite's long-standing survival in the hosts . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [
"medicine"
] | 2013 | Differential Activation of Diverse Glutathione Transferases of Clonorchis sinensis in Response to the Host Bile and Oxidative Stressors |
The methylfolate trap , a metabolic blockage associated with anemia , neural tube defects , Alzheimer’s dementia , cardiovascular diseases , and cancer , was discovered in the 1960s , linking the metabolism of folate , vitamin B12 , methionine and homocysteine . However , the existence or physiological significance of this phenomenon has been unknown in bacteria , which synthesize folate de novo . Here we identify the methylfolate trap as a novel determinant of the bacterial intrinsic death by sulfonamides , antibiotics that block de novo folate synthesis . Genetic mutagenesis , chemical complementation , and metabolomic profiling revealed trap-mediated metabolic imbalances , which induced thymineless death , a phenomenon in which rapidly growing cells succumb to thymine starvation . Restriction of B12 bioavailability , required for preventing trap formation , using an “antivitamin B12” molecule , sensitized intracellular bacteria to sulfonamides . Since boosting the bactericidal activity of sulfonamides through methylfolate trap induction can be achieved in Gram-negative bacteria and mycobacteria , it represents a novel strategy to render these pathogens more susceptible to existing sulfonamides .
Sulfonamides , or SULFA drugs , were the first chemical substances systematically used to treat and prevent bacterial infections [1 , 2] , but the use of these drugs gradually declined because of the emergence of resistant organisms [3] . To increase SULFAs’ potency and prevent further resistance , trimethoprim ( TMP ) , which provides synergy , was later developed [4] . Combination regimens using TMP and SULFAs have effectively treated acute urinary tract infections , bacterial meningitis , Pneumocystis jiroveci pneumonia , and shigellosis , and are commonly used as prophylaxis against recurrent and drug resistant infections [3 , 5 , 6] . Unfortunately , TMP has been the only SULFA booster approved for clinical use , and resistance to both TMP and SULFAs has emerged [7] . In addition , the synergistic effect of TMP remains questionable in many bacteria , including Mycobacterium tuberculosis and Pseudomonas aeruginosa [8 , 9] . To protect the efficacy of SULFAs and safely expand their clinical use [10] , novel SULFA boosters are required . A recent strategy for developing antibiotic boosters is “resisting resistance” [11] , in which inhibitors that suppress resistance mechanisms are used to sensitize host bacteria to antibiotics . Our laboratory recently suggested that targeting antifolate resistance may lead to the development of such adjunctive chemotherapies for SULFAs and TMP [12] . We found that disruption of 5 , 10-methenyltetrahydrofolate synthase ( MTHFS ) , an enzyme responsible for the conversion of N5-formyltetrahydrofolate ( 5-CHO-H4PteGlun ) to N5 , N10-methenyltetrahydrofolate ( 5 , 10-CH+-H4PteGlun ) in the folate-dependent one-carbon metabolic network ( Fig 1A ) , led to severe defects in cellular folate homeostasis thus weakening the intrinsic antifolate resistance in bacteria [12] . TMP and SULFAs are bacteriostatic in minimal media . However , they become more bactericidal in rich media , particularly when cellular levels of glycine , methionine and purines are high . In such conditions , the multifactorial deficiency caused by SULFAs is reduced to a single deficiency of thymine ( Fig 1A , highlighted in red ) , and cells undergoing such “unbalanced growth” succumb to thymineless death [13–17] . Exogenous thymine supplementation reduces the bactericidal activity of SULFAs and TMP [14 , 15] . Classified as folate antagonists , or antifolates , these drugs inhibit bacterial de novo folate biosynthesis ( Fig 1A ) , which is absent in mammalian cells . While SULFAs target dihydropteroate synthase ( DHPS ) , TMP inhibits dihydrofolate reductase ( DHFR ) . Both of these enzymes are required for the formation of folate , a vitamin essential for cell growth across all kingdoms of life . The dominant form of folate in the cell is tetrahydrofolate ( H4PteGlun , with n indicating the number of glutamate moieties ) . This reduced folate molecule functions as a carrier of one-carbon units in multiple metabolic reactions that are required for the production of purines , thymidine , amino acids , and the recycling of homocysteine ( Hcy ) , a non-protein amino acid harmful to long half-life proteins ( Fig 1A ) [18] . Antifolate-mediated folate deficiency affects the biosynthesis of nucleic acids and proteins , as well as other important cellular processes including methylation and homeostasis of Hcy [18] . In humans , defects in Hcy homeostasis , or hyperhomocysteinemia , are often associated with folate and vitamin B12 deficiencies observed in medical conditions such as anemia , neural tube defects , cardiovascular diseases , Alzheimer’s dementia , stroke , cancers , and others [18] . This interconnected metabolic syndrome has been explained by the “methylfolate trap” hypothesis that assigns its cause to defects in the multi-cycling reaction catalyzed by the B12-dependent methionine synthase ( MetH , EC:2 . 1 . 1 . 13 ) ( Fig 1A , highlighted in yellow ) [19–21] . This reaction depends on three components: ( i ) N5-methyltetrahydrofolate ( 5-CH3-H4PteGlun ) , a methyl donor , ( ii ) B12 , the intermediate carrier for the methyl group , and ( iii ) the catalytic activity provided by MetH . Besides the methylation of Hcy to form methionine , this reaction recycles 5-CH3-H4PteGlun back to free H4PteGlun which can be further converted to other folate forms ( Fig 1A ) [20 , 21] . This reaction can be compromised by B12 deficiency and/or mutations affecting MetH enzymatic activity . Consequently , the cellular pool of H4PteGlun is trapped in the methylated form ( 5-CH3-H4PteGlun ) , thus interrupting the normal flow of the one-carbon metabolic network ( Fig 1A ) [21–24] . 5-CH3-H4PteGlun is generated from N5 , N10-methylenetetrahydrofolate ( 5 , 10-CH2-H4PteGlun ) in an upstream reaction catalyzed by methylenetetrahydrofolate reductase ( MTHFR ) , which is irreversible in vivo [25] and suppressed by S-adenosylmethionine ( SAM ) [26] . Since SAM is produced from methionine , inhibition of MetH activity leads to reduced SAM levels , thus resulting in derepression of MTHFR , further accelerating the accumulation of 5-CH3-H4PteGlun and Hcy [26] . Attempts to delete metH in mice were unsuccessful as homozygous knockout embryos all died following implantation [27] . Although it has been studied in humans , and ex vivo in mammalian cells , the existence or physiological significance of the methylfolate trap in bacteria has never been documented . Here we report the identification of the methylfolate trap as a novel determinant of SULFA resistance in bacteria . Upon its formation in response to SULFAs , the methylfolate trap causes impaired homeostasis of folate and related metabolites , including a progressive accumulation of Hcy-thiolactone that is known to be cytotoxic . More importantly , cells undergoing the methylfolate trap are also unable to deplete glycine and nucleotides , and suffer thymineless death induced by SULFAs . This metabolic blockage renders pathogenic bacteria , including M . tuberculosis , P . aeruginosa , Escherichia coli and Salmonella typhimurium more susceptible to existing SULFAs both in vitro and in host macrophages . Furthermore , chemical induction of the methylfolate trap , as shown in our experiments , represents a viable method for boosting the antimicrobial activity of available , clinically approved SULFAs against bacterial pathogens .
A screen of 13 , 500 Himar1-transposon M . smegmatis mutants ( details can be found in Supplemental S1 Text ) identified a collection of strains that displayed normal growth in the absence of antifolates but suffered defects in antifolate resistance . After 2 rounds of drug susceptibility tests , the disrupted genes were mapped using nested PCRs , followed by sequencing . Of the 50 chromosomal loci identified as being responsible for the intrinsic antifolate resistance of M . smegmatis ( S1 Table ) , 31 genes ( 62% ) encoded enzymatic activities , 14 of which ( 28% ) were predicted to be involved in folate metabolism or related pathways . The identification of many genes whose functions are related to folate metabolism indicated that the screen was successful . Overall , the resistance determinants were evenly distributed throughout the M . smegmatis genome with some relatively discrete regions and gaps ( Fig 1B ) . Besides many genes encoding homologs of enzymes of the one-carbon metabolic network and related metabolism of amino acids or nucleotides ( fmt , dcd , gabD , cobIJ , metH , glyA , ygfA , and ygfZ ) , genetic mapping revealed Himar1 insertions in genes that encode proteins previously known to provide non-specific antibiotic resistance ( pknG , mshB , cspB , fbpA , and treS ) [28–33] ( S1 Table ) . In addition , insertions were mapped to chromosomal loci potentially affecting regulatory or signaling processes ( mprA , sigB , sigE , pknG , pafA , pup , pcrB , and pcrA ) , transsulfuration ( cysH and mshB ) , transport ( mmpL and pstC ) , and other cellular activities ( S1 Table ) . Mutants were further profiled using chemical complementation . Para-aminobenzoic acid ( pABA ) or a folate derivative ( Fig 1A , blue rectangles , & Fig 1C ) was added exogenously to support growth in the presence of SULFAs or TMP , which inhibit de novo folate synthesis . These analyses provided useful geno-chemo-phenotypic information to each individual antifolate resistance determinant ( S1 Table ) . Chemical complementation identified a group of SULFA-sensitive , “white” mutants that lost the yellow pigment typically displayed by M . smegmatis ( Fig 1C , marked with asterisks ) . The mutants were unable to use exogenous 5-CH3-H4PteGlu1 to antagonize SULFAs ( Fig 1C , panel ( v ) ) . Genetic mapping showed that four mutants in this subgroup had Himar1 insertions at three different TA dinucleotides within the same gene , msmeg_4185 ( 2xTA499-500 , 1xTA2881-2882 , and 1xTA3091-3092 , S1 Fig ) , which encodes a homolog of B12-dependent methionine synthase . Two other mutants had insertions at TA112-113 of msmeg_3873 , which encodes an enzyme ( CobIJ ) that catalyzes two methylation steps , precorrin-2 C20 methyltransferase [CobI , EC:2 . 1 . 1 . 130] and precorrin-3B C17-methyltransferase [CobJ , EC:2 . 1 . 1 . 131] , of the B12 ( cobalamin ) biosynthetic pathway . Interestingly , the function of these 5-CH3-H4PteGlun-related genes were reminiscent of factors involved in the methylfolate trap , a metabolic disorder thus far only described in mammalian cells ( Fig 2A ) . Whereas the metH-encoded enzyme catalyzes the reaction , cobIJ is required for the de novo biosynthesis of B12 , the cofactor required for MetH activity . Exogenous B12 restored both SULFA resistance and 5-CH3-H4PteGlu1 utilization to cobIJ , but failed to restore the metH strains ( Fig 2B ) , resembling the “pseudo-folate deficiency” phenomenon previously observed in anemia patients ( described in the Discussion ) [19] . To detect the methylfolate trap at a metabolic level , M . smegmatis strains growing in a liquid medium were challenged with sulfachloropyridazine ( SCP ) for half an hour to starve the cells from newly synthesized folate . Cultures were immediately harvested and total folate was extracted in subdued light . Samples added with internal standards were analyzed by LC-MS/MS as previously described [12] . Both metH and cobIJ exhibited 5-CH3-H4PteGlun accumulation compared to wild type M . smegmatis ( Fig 2C ) . Exogenous B12 significantly reduced 5-CH3-H4PteGlun accumulation in the cobIJ mutant , though not to the level of wild type ( Fig 2C ) . This B12-responsive alteration in the cellular folate pool of cobIJ explained its pseudo-folate deficiency-like behavior in susceptibility tests ( Fig 2B ) . In the cobIJ mutant , the metH gene remained intact but its encoded protein did not have enough B12 , due to the Himar1 insertion into cobIJ disrupting de novo B12 biosynthesis , to activate its methionine synthase activity . When B12 was exogenously supplemented , the cofactor activated MetH activity , thus bypassing the B12 synthetic defect allowing for the release of the methylfolate trap . To confirm that MetH and CobIJ contribute to the intrinsic SULFA resistance , and 5-CH3-H4PteGlun metabolism , the encoding genes , msmeg_4185 and msmeg_3873 , respectively , were individually deleted by homologous recombination [34] . Similar to the transposon mutants , the targeted null mutants , MsΔmetH and MsΔcobIJ , displayed increased SULFA susceptibility and impaired utilization of exogenous 5-CH3-H4PteGlu1 whereas in trans expression of metH and cobIJ , respectively , restored both phenotypes ( Table 1 , Fig 2D ) . Exogenous B12 restored both SULFA resistance and 5-CH3-H4PteGlu1 utilization to MsΔcobIJ , but failed to do so for MsΔmetH . Although the mutants were hypersusceptible to all SULFAs tested ( S2 Fig ) , resistance to non-antifolate antibiotics remained unaffected ( S3 Fig ) . While M . smegmatis encodes a B12-independent methionine synthase ( MetE , EC: 2 . 1 . 1 . 14 ) [35] , deletion of metE did not affect SULFA sensitivity ( S4 Fig and S5 Fig ) . These observations confirmed that MetH is essential for normal 5-CH3-H4PteGlun metabolism , which is required for the intrinsic SULFA resistance in M . smegmatis . Mutants lacking metH or cobIJ genes were first constructed from the M . tuberculosis laboratory strain H37Rv ( see S1 Text ) . Sensitivity tests using the MTT method were performed with two minimal media , 7H9-S or Dubos , in the absence or presence of exogenous B12 ( tested range: 1 μM—0 . 3 mM ) . In the absence of B12 , SULFA susceptibility of the H37Rv-derived strains were similar . However , with B12 supplementation , significant differences in SULFA resistance among strains were observed ( Table 1 , Fig 3A ) . While RvΔmetH displayed high susceptibility to sulfamethoxazole ( SMZ ) , the sensitivity level of RvΔcobIJ was unchanged compared to wild type ( Table 1 , Fig 3A ) . In trans expression of metH completely restored wild type SULFA resistance to RvΔmetH ( Table 1 , Fig 3A ) . These results indicated that the methylfolate trap was able to sensitize M . tuberculosis H37Rv to SULFA drugs . Such trap formation , however , requires the absence of methionine synthase activities . In agreement with previous studies [36 , 37] , our data suggested that H37Rv is unable to synthesize B12 de novo , and that this organism relies on its uptake system for obtaining B12 from the environment . In the complete absence of B12 , H37Rv employed the B12-independent methionine synthase MetE to prevent the methylfolate trap . When B12 was added exogenously , MetE activity was inhibited , making RvΔmetH completely null of methionine synthases . In such a condition , the methylfolate trap was formed sensitizing RvΔmetH to SULFA drugs . It is important to note that exogenous supplementation of methionine only partially enhanced SMZ resistance of RvΔmetH ( Table 1 ) , indicating that the lack of methionine due to defective methionine synthases [37 , 38] is not the sole contributor to the enhanced SULFA susceptibility . To further characterize the methionine-unrelated methylfolate trap-mediated SULFA sensitivity , survival of the M . tuberculosis strains treated with SMZ , B12 , and methionine were assayed by serial dilution and colony forming unit ( c . f . u . ) counting . With similar inputs , the survival of RvΔmetH was 3 log10 lower than that of wild type M . tuberculosis H37Rv and the RvΔcobIJ mutant ( Fig 3B ) . This result not only confirmed our observation from the growth inhibition assays ( Table 1 , Fig 3A ) , but further suggested that the methylfolate trap may induce the intrinsic bactericidal activity of SULFA drugs . To further characterize the methylfolate trap in M . tuberculosis , we used CDC1551 , a clinical strain isolated in a 1994–1996 tuberculosis outbreak in the United States [39] , for constructing several strains related to methylfolate trap formation ( S2 Table ) . CDC1551 is a natural metH deletion mutant due to a 1 , 196-bp truncation located at the 3’-terminus of its encoding gene ( mt2183 ) ( Fig 3C ) [38 , 40] . Similar to the M . smegmatis methylfolate trap mutants , colonies of CDC1551 displayed a “white” morphology , differing from the yellow appearance of H37Rv , which resembles wild type M . smegmatis ( S6 Fig ) . To better understand the molecular mechanisms affecting trap formation , SULFA sensitivity tests were performed with a minimal medium ( Dubos ) and a gradient of increasing B12 concentrations ( Fig 3D ) . In the absence of exogenous B12 , CDC1551 ( numbered 2 ) displayed higher SULFA sensitivity compared to the CDC1551 strain in trans expressing the intact metH gene from H37Rv ( CDC1551/metH , numbered 1 ) , indicating that , unlike H37Rv , the B12 biosynthesis is functional in CDC1551 ( Fig 3D ) . The level of internally synthesized B12 was likely enough to partially repress the expression of metE and to activate MetH activity ( see Discussion ) . When cobIJ was deleted ( CDCΔcobIJ/metH and CDCΔcobIJ , numbered 3 and 4 respectively ) , SULFA resistance increased ( Fig 3D ) , possibly due to the derepression of metE in the complete absence of B12 ( similar to H37Rv in minimal medium ) . Deletion of bacA ( numbered 5 and 6 ) , encoding the B12 uptake system in M . tuberculosis [37] , did not have any effect in this condition ( far left panel ) . In the presence of as low as 0 . 25 μM B12 , metE expression appeared to be further suppressed , making CDC1551 highly susceptible to SMZ compared to CDC1551/metH ( Fig 3D , second panel from left ) . The higher the concentration of exogenous B12 , the less SULFA resistance was displayed by CDCΔcobIJ , most likely due to increased suppression of metE . This was not seen in the case of CDCΔcobIJ/metH since MetH was further activated in the presence of B12 , thus compensating for metE suppression . Unlike CDC1551 , CDCΔbacA did not show a severe reduction in SULFA resistance when B12 was added due to its lack of B12 uptake activity . Similarly , but conversely , CDCΔbacA/metH did not show an increased SULFA resistance compared to CDC1551/metH in response to exogenous B12 . Most importantly , as seen with the H37Rv background ( Fig 3A ) , exogenous methionine did not enhance the SULFA resistance of CDC1551-derived strains ( Fig 3D ) . Previous studies suggested that M . tuberculosis is able to uptake and metabolize B12 from its host [41] . To evaluate if the methylfolate trap can form thus affecting the SULFA sensitivity of M . tuberculosis residing within macrophages , strains were used to infect the macrophage cell line J774 . A1 , grown in a medium containing 10% fetal bovine serum . The infected macrophages were treated with SMZ , followed by serial plating of the intracellular bacteria and c . f . u . counting . In both the H37Rv ( Fig 3E ) and the CDC1551 backgrounds ( Fig 3F ) , strains lacking metH exhibited significantly increased sensitivity to SULFA treatment . In trans expression of H37Rv metH ( rv2124c ) restored SULFA resistance to both RvΔmetH and CDC1551 ( Fig 3E and 3F ) . As previously suggested [39] , the proliferation of CDC1551 in macrophages in the absence of SMZ was much faster compared to H37Rv ( Fig 3F ) . However , its survival was more severely reduced compared to H37Rv when the infected macrophages were treated with SMZ ( Fig 3F ) . This enhanced bactericidal activity of SMZ against CDC1551 was reduced in CDC1551/metH , confirming the correlation of MetH activity and the intrinsic resistance of CDC1551 to SULFAs . Together , these results demonstrated that ( i ) the methylfolate trap , when successfully formed , can sensitize M . tuberculosis to SULFAs both in vitro and during infection of host macrophages , ( ii ) the methylfolate trap promotes the bactericidal activity of SULFA drugs , ( iii ) because of its non-functional B12 biosynthetic pathway , H37Rv relies on its uptake system to obtain exogenous B12 , ( iv ) trace amounts of B12 are sufficient to suppress metE expression giving metH a more important role in preventing methylfolate trap formation , and ( v ) because of its truncated metH gene , CDC1551 is intrinsically more susceptible to methylfolate trap formation , rendering it more sensitive to SULFAs both in vitro and during macrophage infection [42] ( Fig 3D and 3F ) . Our laboratory is currently investigating how mutations in metH and genes involved in B12 biosynthesis affect SULFA sensitivity among M . tuberculosis clinical isolates . To assess if the methylfolate trap plays a similar role in SULFA sensitivity in Gram-negative bacteria , we investigated its role in a selected group of significant pathogens with distinct metabolic capacities . Similar to the M . tuberculosis H37Rv strain , E . coli does not synthesize B12 , instead it imports the vitamin via the transport system BtuBCED [43 , 44] . Whereas mutations in btuC , btuE , and/or btuD partially reduce uptake , mutations in btuB completely abolish B12 transport [45] . On a complex medium , an E . coli ΔbtuCED ( b1711 , b1710 , and b1709 , respectively ) triple mutant remained SULFA resistant , whereas ΔmetH ( b4019 ) , ΔbtuB ( b3966 ) , and a ΔbtuBΔCED quadruple mutant all became hypersusceptible ( Fig 4A , Table 1 ) . In serial dilution-spot tests using 125 μg/ml SMZ , these mutants displayed >104 times increased susceptibility compared to wild type BW25113 ( Fig 4A ) . Exogenous B12 was unable to restore SMZ resistance in these mutants due to the absence of MetH or B12 transport activity ( Fig 4A ) . The increased SULFA sensitivity was verified by measuring minimal inhibitory concentrations ( MIC , Table 1 ) , which is defined as the lowest concentration of an antibiotic that inhibits the visible growth of bacteria . To demonstrate methylfolate trap formation at the metabolic level , E . coli cultures were treated with SMZ and total folate was immediately extracted and analyzed by LC-MS/MS [12] . As shown in Fig 4B , 5-CH3-H4PteGlun markedly accumulated in ΔmetH and ΔbtuB compared to the parental strain , confirming methylfolate trap formation . Because of its inability to synthesize B12 de novo , E . coli relies entirely on import to prevent the methylfolate trap . P . aeruginosa is capable of not only synthesizing de novo but also importing B12 from the environment . Transposon mutants with insertions in genes encoding metH ( PA1843 ) , cobI ( PA2904 ) , cobJ ( PA2903 ) , cobH ( PA2905 ) and btuB ( PA1271 ) were obtained from the Pseudomonas Transposon Mutant Collection ( Manoil Laboratory , University of Washington Genome Sciences ) [46] ( S2 Table ) . The mutants were subjected to antifolate susceptibility tests , followed by folate analysis as described above . All P . aeruginosa mutants became more susceptible to SULFA drugs on a complex medium ( Fig 4C , Table 1 ) . The P . aeruginosa metH and btuB mutants displayed identical , and the most severe susceptibility to SULFAs . These strains were at least 105 times more susceptible than wild type as revealed by serial dilution-spotting assays using 125 μg/ml SMZ ( Fig 4C ) . cob mutants were less susceptible compared to these two strains , suggesting that B12 import is more important than de novo synthesis in the condition tested ( Fig 4C , Table 1 ) . Indeed , exogenous B12 reinstated growth of the cob mutants but failed to do the same for metH and btuB ( Fig 4C ) . Chemical analyses also revealed accumulation of the methylfolate trap marker , 5-CH3-H4PteGlun , in both metH and btuB ( Fig 4D ) . Similar experiments with S . typhimurium strains ( John Roth Laboratory , UC Davis , S2 Table ) confirmed the correlation of the methylfolate trap and increased SULFA susceptibility in bacteria ( Table 1 , S7 Fig , and further studies below ) . Similar to M . smegmatis and other Gram-negative bacteria , the deletion of metH , but not metE , resulted in the methylfolate trap and reduced SULFA resistance in S . typhimurium on complex media ( S7 Fig ) . The absence of metH , hence the methylfolate trap , led to increased susceptibility to SULFA drugs classified in all categories ( Fig 5A ) , but not to folate-unrelated antibiotics ( S8 Fig ) . To investigate if the effect of the methylfolate trap was bactericidal or bacteriostatic , S . typhimurium metH ( + ) and metH ( - ) strains were spotted on filters ( ~104 cells/filter ) , which were placed on the surface of Luria-Bertani ( LB ) agar plates supplemented with or without SMZ . Following 24 h of incubation at 37°C , cells from the inoculated filters were resuspended , and colony forming units ( c . f . u . ) were measured by serial dilution and plating . On LB agar free of SULFA , both metH ( + ) and metH ( - ) proliferated to 106 times more cells than the input ( Fig 5B ) . In the presence of 125 μg/ml SMZ , growth of metH ( + ) was normal whereas only 0–8 . 5% of the metH ( - ) input survived ( Fig 5B ) , indicating an enhanced bactericidal effect of SMZ due to the methylfolate trap . In liquid LB , addition of 2 . 5 mg/ml SMZ similarly reduced growth of metH ( - ) while still allowing growth of metH ( + ) ( Fig 5C ) , confirming the correlation between SULFA resistance and MetH activity . To investigate if the increased susceptibility was due to enhanced import , the SULFA uptake of S . typhimurium strains was measured using radioactive SMZ . However , both metH ( + ) and metH ( - ) displayed identical uptake following the addition of SMZ to the medium ( S9 Fig , panel A ) . We next examined the effect of the methylfolate trap on the synthesis of macromolecules ( DNA , RNA and protein ) during SULFA treatment . Cells of metH ( + ) or metH ( - ) bacteria , growing in the presence of SMZ , were labeled using [3H]-thymidine , [3H]-uracil , or [35S]-methionine , respectively . While DNA and protein synthesis were not affected by the methylfolate trap during SULFA treatment , RNA synthesis was significantly reduced in cells suffering the metabolic blockage ( S9 Fig , panels B-D ) . To assess changes in the folate pool during SULFA-induced methylfolate trap formation , S . typhimurium cells growing in liquid LB medium were treated with SMZ , followed by sample collection . Folate was extracted and individual species quantified using LC-MS/MS . In the presence of MetH , combined levels of both methylated ( 5-CH3-H4PteGlun ) and non-methylated folate species ( R-H4PteGlun R ≠ CH3 ) immediately and continuously declined in response to SMZ ( Fig 5D , top and middle panels , red bars; see also S10 Fig for the dynamics of individual species ) . In contrast , in metH ( - ) cells , 5-CH3-H4PteGlun gradually accumulated following SMZ treatment ( Fig 5D , top panel , blue bars ) . Levels of non-methylated folate species in metH ( - ) gradually declined for the first hour , then remained constant for the remainder of the experiment ( Fig 5D , middle panel , blue bars ) . This result indicated possible cellular feedback , either through an increase in de novo H4PteGlun synthesis or rearrangement in the inter-conversion network of one-carbon metabolism . To further analyze metabolic alterations in response to such folate homeostatic defects , post-SMZ treatment levels of 41 metabolites were profiled using LC-MS/MS-based metabolomics . Cells were sampled from growth curves similar to those in Fig 5C from which metabolites were extracted and analyzed by the Metabolomics Lab at the Roy J . Carver Biotechnology Center ( University of Illinois at Urbana-Champaign ) . Metabolic abnormalities caused by the SMZ-induced methylfolate trap include the accumulation of intermediates within the methionine-homocysteine cycle ( Figs 5E and 6A , orange ) , glycine ( Figs 5E and 6B , red ) and nucleotides ( Figs 5E and 6C , purple ) , as discussed in more detail below . The MetH reaction connects the one-carbon metabolic network with the methionine cycle through its conversion ( methylation ) of Hcy to methionine ( S9 Fig , panel E ) . Therefore , impaired MetH would lead to the accumulation of not only 5-CH3-H4PteGlun , but also Hcy , causing hyperhomocysteinemia . In the cell , Hcy is further converted to Hcy-thiolactone , which is cytotoxic due to its interaction with physiologically important proteins [47 , 48] . Because it is neutral at physiological pH ( pKa = 6 . 67 ) , Hcy-thiolactone is steadily secreted into exogenous media following its production from Hcy [48] . Besides harvesting cells for folate and metabolomic analyses ( Fig 5D and 5E ) , culture filtrates from metH ( + ) and metH ( - ) growing in the presence of SMZ were also collected for Hcy-thiolactone analysis ( S1 Text ) [49] . As shown in Fig 6A , cells of metH ( - ) accumulated S-adenosylhomocysteine ( SAH ) , which led to higher levels of Hcy-thiolactone in the medium compared to metH ( + ) ( S9 Fig , panel F ) . In the presence of MetH ( red circle ) , production of methionine ( Fig 6A ) and glycine ( Fig 6B ) rapidly dropped while levels of nucleotides ( Fig 6C ) including aminoimidazole carboxamide ribonucleotide ( AICAR ) , a precursor of purine synthesis , slightly increased during the first half an hour to one hour of SMZ treatment . Thereafter , synthesis of methionine and glycine resumed but nucleotides underwent continuous depletion . In the absence of MetH ( blue triangle ) , methionine synthesis slightly increased ( Fig 6A ) , most likely due to increased uptake , nucleotides levels also increased ( Fig 6C ) , but glycine levels slightly declined ( Fig 6B ) in the first hour . After this time period , nucleotides , especially dUMP , remained highly elevated , methionine levels declined and remained constant while glycine levels increased and remained elevated . Antifolate-responsive depletion of intracellular glycine and purines was recently proposed as an E . coli mechanism to escape thymineless death [15] . To test if thymine plays a role in the methylfolate trap-promoted bactericidal activity of SULFA , this nucleotide precursor was added to medium and the survival of strains was evaluated by serial dilution and plating method . Interestingly , thymine abolished the SULFA-induced death of the metH ( - ) strain , and restored its growth ( Fig 6D ) . These results suggest that the methylfolate trap promotes the intrinsic thymineless death of bacteria by SULFA drugs , by causing an imbalance in nucleotide levels while preventing cellular depletion of glycine . To investigate if the methylfolate trap renders bacteria more susceptible to SULFAs in a host cell environment , we first monitored the intracellular survival of S . typhimurium strains in J774A . 1 , a macrophage cell line commonly used for antibiotic sensitivity testing [50] . When the infected macrophages were treated with SMZ at a concentration sub-inhibitory for the S . typhimurium parental strain , mutants undergoing the methylfolate trap displayed significant defects in survival ( Fig 7A ) . The survival of the S . typhimurium strains in macrophages resembled the patterns of in vitro sensitivity ( S7 Fig ) , suggesting a similar role of the methylfolate trap in promoting SULFA susceptibility of intracellular bacteria . To assess if SULFA susceptibility of the intracellular bacteria can be promoted through pharmacological induction of the methylfolate trap , we sought to restrict B12 bioavailability using a chemical approach ( Fig 7B ) . The cellular uptake and conversion of exogenous B12 ( cyanocobalamin ) to biologically active cofactors ( adenosylcobalamin and methylcobalamin ) in mammalian cells requires the enzymatic activity of CblC , also known as MMACHC ( for methylmalonic aciduria ( cobalamin deficiency ) cblC type , with homocystinuria ) [51] . To investigate if B12 bioavailability , hence SULFA sensitivity , of intracellular S . typhimurium could be controlled through CblC inhibition , expression of cblC in macrophages THP-1 was depleted using RNA interference . Transfection with cblC-specific siRNA effectively reduced CblC expression , detected by Western Blot using a CblC monoclonal antibody ( Fig 7C , top panel ) . The reduced cblC expression was found to correlate with increased B12 starvation of the intracellular S . typhimurium bacillus as detected by a B12 molecular probe ( Fig 7C , middle ) [52] . Within such CblC-depleted macrophages , S . typhimurium became more SMZ susceptible as determined by c . f . u plating assays ( Fig 7C , bottom ) . We recently developed Coβ-4-ethylphenylcob- ( III ) alamin ( EtPhCbl ) [53] , a cobalamin analog that can function as a vitamin B12 antagonist ( or “antivitamin B12” ) [53 , 54] . EtPhCbl effectively binds to CblC but resists dissociation from the protein , thereby blocking CblC from its normal functions of decyanation and dealkylation of newly internalized cyanocobalamin and methylcobalamin , respectively [55 , 56] . Because bacteria do not have CblC homologs , EtPhCbl had no effect when used directly on bacterial cells ( S11 Fig ) . To test whether EtPhCbl increases methylfolate trap-mediated SULFA susceptibility in bacteria residing within host cells , macrophages were first infected with S . typhimurium . Thereafter , the infected cells were treated with SMZ , EtPhCbl , or their combination . Cells were then lysed and intracellular bacteria determined by c . f . u . plating assays . Whereas SMZ or EtPhCbl alone did not affect the intracellular survival of S . typhimurium , their combination resulted in both B12 starvation ( Fig 7D , top and middle ) , and a significant c . f . u . reduction ( Fig 7D , bottom ) to the intracellular bacillus .
We first constructed a large library of transposon insertion mutants in M . smegmatis . The size of the library was approximately 2 times the number of genes in the M . smegmatis genome ( 6 , 717 protein-coding genes and 54 RNA-coding genes , http://www . genome . jp/kegg-bin/show_organism ? org=msm ) . Screening this library , we identified 50 chromosomal loci responsible for the intrinsic antifolate resistance in M . smegmatis ( S1 Table ) . Further investigation of the inserted genes revealed many novel pathways previously unknown to be involved in bacterial intrinsic antifolate resistance . For example , we recently reported the role of 5 , 10-methenyltetrahydrofolate synthase ( MTHFS , encoded by msmeg_5472 ) , which converts 5-CHO-H4PteGlun , a proposed storage form of folate , to 5 , 10-CH+-H4PteGlun , in cellular folate homeostasis and bacterial antifolate resistance [12] . These studies , including the current work reported in this paper , confirm the richness of potential drug targets in this pathway as previously postulated [57 , 58] . The fact that many loci were repeatedly identified in the screen confirmed the saturation of the library . It is however important to note that our screening procedures only selected mutants that showed “normal growth” in the absence of antifolates; thus resistance determinants encoded by essential genes or genes whose mutation affected M . smegmatis growth on NE medium in the absence of antifolates were not included in this pool . In further studies of the mutant library , we have now discovered another novel mechanism of intrinsic SULFA resistance in bacteria referred to as the methylfolate trap , which occurs when cellular H4PteGlun is trapped in a single methylated form , 5-CH3-H4PteGlun [21 , 59] . We show that the methylfolate trap increases the bactericidal activity of SULFA drugs against mycobacteria and Gram-negative bacteria . The methylfolate trap hypothesis was first proposed by Herbert and Zalusky in 1962 to explain the cause of megaloblastic anemia observed in patients deficient in folate and vitamin B12 [19] . Besides the typical low blood count and macrocytosis , cells from those patients encountered a “pseudo-folate deficient” state , in which folic acid injected into tissues rapidly disappeared while 5-CH3-H4PteGlun “piled up” in sera . However , simultaneous treatment of many of those patients with vitamin B12 immediately corrected the folate leakage and blood count normalized [20 , 59 , 60] . These phenomena were hypothesized to be a result of deficiencies in the B12-dependent methionine synthase ( MetH ) activity , which converts 5-CH3-H4PteGlun and Hcy to H4PteGlun and methionine , respectively . This hypothesis was supported by the fact that all patients with inborn genetic errors in the metH gene suffer from anemia or developmental delay; and exhibit accumulation of 5-CH3-H4PteGlun and Hcy [61 , 62] . However , direct genetic evidence connecting metH and the methylfolate trap has not been established because constructing metH knockout mice has proved unsuccessful thus far [27] . Nonetheless , the methylfolate trap hypothesis is now widely accepted to explain the relationships of B12 , folate , and Hcy homeostasis in many human diseases [63] . The methylfolate trap and its physiological consequences have never been described in bacteria , plants or microbial eukaryotes , possibly because these organisms are able to synthesize folate de novo , thus minimizing the trap’s effects . Interestingly , our data show that the methylfolate trap is lethal to bacteria when it is formed in the presence of SULFA drugs , which inhibit de novo folate biosynthesis . Due to the lack of de novo folate synthesis , mammalian cells undergoing the methylfolate trap exhibit a depletion of non-methyl folate species , consequently leading to reduced synthesis of amino acids and nucleotides from the one-carbon metabolic network . By contrast , the levels of non-methyl folate species in bacterial cells experiencing the trap only modestly reduced or did not change , while total folate elevated because of the increase in 5-CH3-H4PteGlun levels ( Figs 2C , 4B , 4D and 5D ) . This was most likely due to an increase in de novo folate synthesis in response to the continuous loss of folate molecules trapped in the irreversible 5-CH3-H4PteGlun form . Such a response leads to two possible lethal consequences: ( i ) a wasteful cycle of synthesis and loss of H4PteGlun which rapidly depletes cellular resources , or ( ii ) an uncoordinated increase in activity of the early steps preceding the MetH reaction in the one-carbon metabolic network ( Fig 1A ) . Because thymidylate synthase is a rate-limiting reaction , such an increase in H4PteGlun influx in the absence of MetH would lead to increased synthesis of some amino acids and nucleotides while levels of thymidine nucleotides remain low , thus promoting “unbalanced” growth that causes thymineless death [13] . Our metabolomic data shed light on these possibilities . Besides the extracellular accumulation of Hcy-thiolactone ( S9 Fig , panel F ) , which may be deleterious to exogenously functioning molecules , cells undergoing the methylfolate trap were unable to deplete glycine and nucleotides ( Fig 6B and 6C ) . Cellular depletion of glycine and purines was found necessary for bacterial escape from thymineless death , a known contributor to the bactericidal activity of antifolates [15 , 16] . Although thymidine triphosphate ( dTTP ) was not detectable in cells subjected to our experimental conditions , the level of deoxyuridine monophosphate ( dUMP ) , a precursor of dTTP , increased 700 fold in the absence of MetH after 8 hours of SMZ treatment ( Fig 6C , 219 . 26 in metH ( - ) versus 0 . 31 in metH ( + ) , p = 0 . 0371 ) , indicating low activities of thymidylate synthase ( TS , Fig 1A ) in the presence of the methylfolate trap . Cellular accumulation of dUMP , leading to robust dUTP production , has been known to contribute to thymineless death by causing misincorporation of uracil into DNA [64] . Importantly , exogenous supplementation of thymine completely abolished the SULFA-induced death in metH ( - ) ( Fig 6D ) . In addition , cells suffering the methylfolate trap displayed unchanged synthesis of proteins and DNA but reduced synthesis of RNA ( S9 Fig , panel C ) , a hallmark exhibited by bacterial cells that undergo thymineless death [65 , 66] . Together , our studies suggest that the methylfolate trap boosts the bactericidal activity of SULFAs by inducing thymineless death . It is important to note that many bacteria also encode MetE , a B12-independent methionine synthase [35] . However , catalytic activity of MetE is more than a hundred fold lower than that of MetH [67 , 68] , and the expression of metE is sensitive to B12 exposure [38] , making MetH the dominant methionine synthase . In fact , our data indicated that neither deletion nor overexpression of metE affected SULFA susceptibility in M . smegmatis ( S4 Fig and S5 Fig ) and S . typhimurium ( S7 Fig ) , and that de novo synthesized B12 contributes to partially inhibiting metE expression in autotrophic bacteria . In the complete absence of exogenous B12 in minimal media , B12 auxotrophic bacteria such as the M . tuberculosis laboratory strain H37Rv are able to use MetE activity to prevent trap formation . However , exposure to minute amounts of B12 is enough to suppress metE expression . With previous studies showing that functionally adequate levels of B12 are accessible to bacterial pathogens during vertebrate host infections [41] , the role of MetE in the methylfolate trap-mediated SULFA sensitization is likely negligible . The fact that metH deletion leads to increased SULFA sensitivity in H37Rv during macrophage infection ( Fig 3E ) further suggested that this bacterium is able to acquire B12 from the host cell , and that the acquired B12 is sufficient for preventing methylfolate trap formation . Similar to mammalian cells , bacteria undergoing restricted de novo folate synthesis caused by SULFAs relied on vitamin B12 for preventing methylfolate trap formation . Accordingly , reduced B12 bioavailability could sensitize some bacterial pathogens to SULFAs . Our experiments presented in Fig 7 provide a proof-of-concept that this folate antagonistic strategy , namely the chemical promotion of the methylfolate trap , is feasible for inducing the killing of pathogenic bacteria by SULFAs . However , targeting B12 bioavailability by general antivitamin B12 molecules may not be effective for some bacteria , providing the heterogeneity of B12 biosynthesis and uptake . In addition , it is currently not known if such antivitamin B12 compounds play a role in the regulation of B12 synthesis or uptake in the targeted bacterial pathogens . Another challenge is how to develop methylfolate trap inducers that are specific for bacteria , thus causing no significant toxicity to mammalian cells . In this regard , targeting bacterial proteins involved in B12 uptake and salvage , which are distinct from those of the mammalian counterparts , may provide a possible strategy . As we have previously proposed [12 , 57 , 58] , antifolate resistance determinants such as the methylfolate trap represent potential targets for the development of SULFA boosters , which not only protect the efficacy of SULFAs but also increase their potency against drug resistant pathogens . With the increasing use of co-trimoxazole ( SMZ plus TMP ) in prophylactic treatments of HIV positive patients throughout the world [10] , such SULFA boosters are urgently needed .
Strains , plasmids , and primers used in this study are listed in S2 , S3 and S4 Tables of the Supporting Information , which also contain information on the genetic screen and identification of antifolate-sensitive mutants , targeted gene deletion , genetic and chemical complementation , extraction and analysis of cellular folate derivatives , and antibiotic susceptibility tests ( S1 Text ) . M . smegmatis mc2155 and its derived transposon mutants were grown in LB broth or 7H9 ( Difco ) supplemented with glucose and 0 . 5% Tween 80 . M . tuberculosis strains were grown in 7H10-OADC or Dubos-ADC media ( Difco ) . Unless otherwise stated , Gram-negative bacteria were grown in LB broth or LB agar . Statistical analyses were conducted using GraphPad Prism 5 . 0f software ( La Jolla , CA ) . Students two-tailed t-test was used to analyze the statistical significance of differences between groups . Other methods used in this study can be found in the Supporting Information ( S1 Text ) . | Sulfonamides were the first agents to successfully treat bacterial infections , but their use later declined due to the emergence of resistant organisms . Restoration of these drugs may be achieved through inactivation of molecular mechanisms responsible for resistance . A chemo-genomic screen first identified 50 chromosomal loci representing the whole-genome antifolate resistance determinants in Mycobacterium smegmatis . Interestingly , many determinants resembled components of the methylfolate trap , a metabolic blockage exclusively described in mammalian cells . Targeted mutagenesis , genetic and chemical complementation , followed by chemical analyses established the methylfolate trap as a novel mechanism of sulfonamide sensitivity , ubiquitously present in mycobacteria and Gram-negative bacterial pathogens . Furthermore , metabolomic analyses revealed trap-mediated interruptions in folate and related metabolic pathways . These metabolic imbalances induced thymineless death , which was reversible with exogenous thymine supplementation . Chemical restriction of vitamin B12 , an important molecule required for prevention of the methylfolate trap , sensitized intracellular bacteria to sulfonamides . Thus , pharmaceutical promotion of the methylfolate trap represents a novel folate antagonistic strategy to render pathogenic bacteria more susceptible to available , clinically approved sulfonamides . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"blood",
"cells",
"antimicrobials",
"cell",
"physiology",
"medicine",
"and",
"health",
"sciences",
"immune",
"cells",
"pathology",
"and",
"laboratory",
"medicine",
"chemical",
"compounds",
"pathogens",
"drugs",
"immunology",
"microbiology",
"cell",
"metabolism",
"organic",
"compounds",
"bacterial",
"diseases",
"antibiotics",
"enterobacteriaceae",
"amino",
"acids",
"pharmacology",
"bacteria",
"drug",
"metabolism",
"bacterial",
"pathogens",
"salmonella",
"typhimurium",
"infectious",
"diseases",
"white",
"blood",
"cells",
"animal",
"cells",
"proteins",
"medical",
"microbiology",
"microbial",
"pathogens",
"chemistry",
"actinobacteria",
"methionine",
"salmonella",
"pharmacokinetics",
"sulfur",
"containing",
"amino",
"acids",
"biochemistry",
"organic",
"chemistry",
"cell",
"biology",
"mycobacterium",
"tuberculosis",
"microbial",
"control",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"physical",
"sciences",
"macrophages",
"organisms"
] | 2016 | Methylfolate Trap Promotes Bacterial Thymineless Death by Sulfa Drugs |
DNA double-strand break ( DSB ) repair occurring in repeated DNA sequences often leads to the generation of chromosomal rearrangements . Homologous recombination normally ensures a faithful repair of DSBs through a mechanism that transfers the genetic information of an intact donor template to the broken molecule . When only one DSB end shares homology to the donor template , conventional gene conversion fails to occur and repair can be channeled to a recombination-dependent replication pathway termed break-induced replication ( BIR ) , which is prone to produce chromosome non-reciprocal translocations ( NRTs ) , a classical feature of numerous human cancers . Using a newly designed substrate for the analysis of DSB–induced chromosomal translocations , we show that Mus81 and Yen1 structure-selective endonucleases ( SSEs ) promote BIR , thus causing NRTs . We propose that Mus81 and Yen1 are recruited at the strand invasion intermediate to allow the establishment of a replication fork , which is required to complete BIR . Replication template switching during BIR , a feature of this pathway , engenders complex chromosomal rearrangements when using repeated DNA sequences dispersed over the genome . We demonstrate here that Mus81 and Yen1 , together with Slx4 , also promote template switching during BIR . Altogether , our study provides evidence for a role of SSEs at multiple steps during BIR , thus participating in the destabilization of the genome by generating complex chromosomal rearrangements .
The maintenance of genome integrity is crucial to prevent cell death in all organisms . Chromosomal rearrangements such as reciprocal translocations , deletions , inversions and duplications threaten genomic stability and must be avoided to prevent cancer development and genomic disorders [1] . The occurrence of unfaithful repair of DNA double-strand breaks ( DSBs ) is widely admitted to be the main source of chromosomal rearrangements [2]–[5] . Nonhomologous End Joining ( NHEJ ) and Homologous recombination ( HR ) constitute the main pathways of DSB repair . While NHEJ seal the broken DNA ends by simple religation , HR uses sequence homology between the DSB ends and an intact template for repair and is typically considered as error free . Nevertheless , in a number of cases , such as when HR occurs between non-allelic DNA sequences or DNA repeated sequences or HR is used for the repair of DSB ends containing different levels of similarity , irreversible genomic changes can take place . Thus , when only one end of a DSB shares homology with other sequences in the genome , repair by HR can occur through a replication mechanism termed break-induced replication ( BIR ) that often gives rise to non-reciprocal translocations ( NRTs ) [6]–[9] . BIR requires HR canonical factors such as Rad52 and Rad51 to allow efficient strand invasion of the repair template and to form a Displacement-loop ( D-loop ) that can be extended by DNA synthesis from the invading 3′ DSB end [10] , [11] . In the absence of another DSB end to capture the newly synthesized strand or to independently invade the homologous template , the strand invasion intermediate is thought to be converted into a DNA replication fork capable of replicating an entire chromosome arm until encountering a telomere , a centromere , or a converging replication fork [12]–[15] . A key factor in this process is Pol32 , the non-essential subunit of DNA polymerase δ , which is dispensable for normal replication but essential for BIR [13] . Another feature of BIR is the unstable nature of its replication intermediates . It has been shown that DSB repair by BIR can occur through several rounds of strand invasion , synthesis , and dissociation from the invaded template [16] . Within dispersed repeated sequences , template switching during BIR can generate complex chromosomal rearrangements [16]–[21] . BIR reactions can also be aborted to end in half-crossovers [22] , [23] . Half-crossovers cause NRTs , leaving the template that has been used for repair broken . They are similar to the NRTs observed in humans , which are involved in the cascade of genomic instability characteristic of human cancer cells [24] . Little is known about how BIR intermediates are processed to allow the establishment of a replication fork after strand invasion , and to cause template switching and half-crossovers . Eukaryotic cells have evolved a set of DNA structure-selective endonucleases ( SSEs ) that possess different substrate specificity for various DNA branched molecules during HR , such as D-loops , replication forks , flaps and Holliday junctions ( HJs ) . Conversion of a D-loop into a replication fork during BIR may require the endonucleolytic resolution of a single Holliday junction [16] , [25] . In budding yeast , Mus81-Mms4 and Yen1 are the only nuclear enzymes capable of cleaving intact HJs with high efficiency in vitro [26] , [27] . In vivo , their functions seem to overlap during DSB repair [28]–[31] . Notably , Mus81-Mms4 is required for recombination-mediated DNA repair at replication forks [32]–[34] and has been shown to play a role in BIR intermediate processing [23] . Additionally , the Slx1–Slx4 nuclease complex may cleave perturbed replications forks [35]–[37] . Loss of Slx1–Slx4 , as loss of Mus81-Mms4 , increases the number of gross chromosomal rearrangements in yeast [38] . Slx4 also acts independently of the Slx1 catalytic subunit to interact with several other factors . For instance , Slx4 binds to the 3′ flap endonuclease complex Rad1–Rad10 to facilitate the removal of non-homologous tails during HR [39] , [40] . Rad1–Rad10 complex , which functions in nucleotide excision repair ( NER ) as well as in DSB repair , has also been involved in the formation of translocations in yeast through a single-strand annealing ( SSA ) mechanism [41] , [42] . Together , these SSEs form a complex network to ensure genome stability but their specific roles and their interactions at recombination structures remain unclear . In this study , we used a new assay that generates chromosomal rearrangements after DSB repair using dispersed repeated sequences . Chromosomal rearrangements involved multiple rounds of template switching and some events ended in half-crossovers , generating NRTs . We investigated the role of SSEs in the processing of BIR intermediates combining mutations of Mus81 , Rad1 , Yen1 , Slx1 and Slx4 . Our results show that these SSEs act at multiple steps during BIR . First , we uncovered that Mus81 and Yen1 function to allow efficient BIR , thus causing translocations . In the absence of Mus81 , Yen1 and Slx4 , we observed that template switching during BIR decreased significantly . Altogether , our results led to new insights into the BIR mechanism and the functional role of SSEs in chromosomal rearrangements .
We designed an experimental system using DNA repeated sequences dispersed over two yeast chromosomes for the analysis of chromosomal rearrangements induced by a single DSB . We took advantage of the presence of the MAT , HMR and HML loci on chromosome III , involved in mating-type switching . The MAT locus is composed of five regions called W , X , Y , Z1 and Z2 [43] . MATa differs from MATα by the Ya and the Yα sequences , respectively . MATa shares Ya with HMR and MATα shares Yα with HML . Together , MAT , HMR and HML share two homologous regions flanking the Y sequences , termed X and Z1 ( see Figure 1A ) . All three loci contain a cleavage site for the HO endonuclease at the junction between the Y and Z1 regions , but only the MAT locus is susceptible to being cleaved upon HO expression because of the active repression of HMR and HML loci [44] . The strains used in this study harbor a MATa-inc mutation , a G to A substitution at position Z1–2 , which impedes HO cleavage at MAT [45] , [46] . The HO endonuclease gene under control of the GAL1 inducible promoter was integrated at the ADE3 locus and a 240-bp Ya-Z1 fragment containing a HO-cleavable site , along with the URA3 marker , was inserted in the chromosome VII left arm at the ADH4 locus ( Figure 1A ) . Upon galactose addition to the culture medium , HO would cleave this Ya-Z1 fragment asymmetrically into a centromeric 198-bp fragment and a telomeric 42-bp fragment ( Figure 1A ) . To assay DSB repair in this system , we plated cells on a synthetic medium ( SC ) with either 2% glucose or 2% galactose to assay survival after HO expression and successive breakage of chromosome VII . The WT strain exhibited a survival frequency of 82% ( Figure S1 ) , demonstrating efficient DSB repair . We restreaked survivor colonies from galactose-containing plates on glucose medium to repress HO expression . Then , survivors were concomitantly replica-plated on media lacking uracil and on media containing 5-FOA , a drug that generates a toxic metabolite in Ura+ cells , to distinguish between the colonies that maintained or lost the URA3 marker of chromosome VII ( Figure 1B ) . We recollected Ura− 5-FOA-resistant ( 20% ) , Ura+ 5-FOA-sensitive ( 12% ) , and Ura+ 5-FOA-resistant colonies ( 68% ) . The latter contained both Ura+ and Ura− cells and may result from differential repair of two DSBs generated on sister chromatids during the S or G2 phase of the cell cycle . To analyze mixed Ura+ 5-FOA-resistant colonies , we first separated Ura+ cells from 5-FOA-resistant cells by restreaking colonies on media lacking uracil or containing 5-FOA . Further analyses of 5-FOA-resistant and Ura+ cells were performed by pulse-field gel electrophoresis ( PFGE ) to look for chromosomal translocations . Southern analysis using a probe specific to chromosome VII , proximal to the DSB site , and a probe specific to the region between HMR locus and chromosome III telomere ( 7L probe and 3R1 probe , respectively , Figure 1A ) revealed the presence of non-reciprocal translocations ( NRTs ) between chromosome VII and chromosome III in all 5-FOA-resistant cells ( Figure 1C , 1D , and Figure S2A , S2B ) . NRTs were rarely observed in Ura+ cells ( Figure S2B ) . Given a survival rate of 82% , the frequency of translocants ( frequency of 5-FOA-resistant survivors among the whole population that did not undergo DSB induction ) was 44% in the WT strain . We mainly observed two types of NRTs , which contained chromosome III sequences starting from the MAT or HMR loci to the telomere fused to chromosome VII at the break site ( termed T7/3-MAT and T7/3-HMR translocations for more clarity , Figure 1G ) . Interestingly , some survivors lacked linear chromosome III at its expected size in the PFGE analysis ( Figure 1D , 1E ) . When we used a probe specific to the region between MAT and HMR loci ( 3R2 probe , Figure 1A ) , chromosome III was detected in the well , consistent with a particular structure that did not allow it to enter the gel ( Figure 1E ) . One possibility was that chromosome III became circular , as had been observed more than three decades ago [47] , [48] . To test this , we digested chromosomal DNA with AsiSI , which cuts chromosome III at a unique inner location . Indeed , AsiSI digestion released chromosome III into the gel ( Figure 1E ) . Circular chromosomes III have been shown to be the product of recombination between unrepressed HMR and HML loci , which share extensive homology in the X and Z1 regions ( Figure 1A ) [47] . We observed that the size of the AsiSI linearized chromosome was concordant with a chromosome III that would have lost all subtelomeric sequences located beyond the HM loci . To further demonstrate that circular chromosomes III occurred by recombination between HMR and HML , we assayed whether chromosome III circularization would be seen in hmlΔ cells . We observed that , in the absence of HML locus , all survivors contained a linear chromosome III at its expected size ( Figure 1F ) . Chromosome III circularization pushed us to investigate the occurrence of a DSB at HMR . HMR is naturally packed into heterochromatin to repress its expression and also to impede cleavage by HO [44] . We asked if HMR invasion would release its repression and cause HO cleavage . We first assayed HMR cleavage by Southern at several times after induction of chromosome VII cleavage by HO and did not observe any cuts in the WT ( Figure S3C ) . We only detected HMR cleavage in a hmlΔ matΔ strain , in which HMR cleavage represented about a third of chromosome VII cleavage by HO ( 17% versus 46% after 3 h of induction; Figure S3B , S3C ) . Indeed HMR cleavage was dependent on cleavage at chromosome VII since no cuts were observed in the hmlΔ matΔ strain without HO cut site on chromosome VII ( Figure S3D ) . We concluded that the induction of a single DSB by HO in chromosome VII permitted another less efficient DSB in chromosome III at the HMR locus as a secondary event . The efficiency of HMR cleavage is expected to be even less than 17% in the WT strain since the MATa-inc locus is also available as a template for strand invasion in this strain . HMR cleavage likely induced chromosome III circularization , which we observed in 50% of translocants in the WT strain . Apart from T7/3-MAT , T7/3-HMR translocations and chromosome III circularization , we also observed rare types of rearrangements of chromosome VII and III whose nature has not been addressed in this study . Together , these results demonstrate that complex chromosomal rearrangements are occurring at a high frequency in our experimental system , allowing us to investigate the molecular and genetic bases of these events . We performed kinetic experiments to follow the recombination intermediates that gave rise to chromosomal rearrangements using our system . First , a PCR-based assay was used to monitor new DNA synthesis primed from the 3′ end of the invading strand of chromosome VII . Genomic DNA was extracted at different times after HO induction and PCR was made with one primer specific to chromosome VII ( p7 , Figure 2A ) , proximal to the DSB , and primers specific to each potential template used for recombination , MATa-inc and HMR ( p3-M and p3-R , Figure 2A ) . In WT cells , 35 cycles of PCR amplification permitted to detect products that most likely correspond to newly synthesized DNA fragments at both MAT and HMR loci 1 h after DSB induction ( Figure 2B ) . Using more quantitative conditions , we detected the same products after 25 cycles of PCR amplification at 2 h of DSB induction ( Figure S4 ) . The amount of products increased over time . We then monitored the appearance of repair intermediates directly by Southern analysis of genomic DNA digested by EcoRV , using a probe specific to chromosome VII , proximal to the DSB site ( 7L probe ) . We detected recombination intermediates between chromosome VII and chromosome III at HMR ( 7/3-HMR , Figure 2C , left ) , which most likely correspond to newly synthesized DNA fragments 5 h after DSB induction . At MAT , we observed recombination intermediates 24 h after DSB induction ( 7/3-MAT , Figure 2C , left ) . Southern analysis also detected an unexpected band that corresponded in size to chromosome VII fused to chromosome III sequences at the HML locus ( 7/3-HML , Figure 2C , left ) . We confirmed this assumption by re-probing the Southern membrane with a probe specific to HML ( data not shown ) . The 7/3-HML band most likely corresponds to newly synthesized DNA fragments primed from chromosome VII DSB end at HML and appeared concomitantly with 7/3-HMR intermediates 5 h after DSB induction ( Figure 2C ) . We monitored what was most likely new DNA synthesis primed from the 3′ DSB end of chromosome VII invading HML by PCR . We detected intermediates 2 h after DSB induction ( p7/p3-L , Figure 2B , Figure S4 ) . The DSB end of chromosome VII assayed by PCR and Southern only shares sequence homology with MATa-inc and HMR . Hence , the signals detected at HML would be the consequence of a template switching from MATa-inc or HMR to HML after duplication of the Z1 region that is common to the three loci . In a similar way , template switching could occur between MATa-inc and HMR . 7/3-HMR and 7/3-HML intermediates detected by Southern appeared about 3–4 h after detection of priming of the 3′ invading DSB end by PCR ( Figure 2B , 2C ) . This difference could be due to the difference of sensitivity between the two techniques . Alternatively , this could reflect a transition between the elongation of the 3′ invading DSB end , a step that is common to GC and BIR , and the establishment of an active replication fork required for BIR [11] , [14] . Using translocants containing T7/3-MAT and T7/3-HMR translocations , we confirmed the size of the repair intermediates detected by Southern ( Figure 2C , right ) . We did not recover any translocant containing translocations corresponding to repair intermediates detected at HML . Together , these data show that one unique feature of our experimental system is that it allows the detection of events of template switching between MATa-inc and HML and possibly between MATa-inc and HMR during chromosomal rearrangement . These events of template switching likely participated to give rise to the formation of T7/3-MAT and T7/3-HMR translocations and dicentric chromosomes resulting from BIR completed from HML ( Figure 2D ) . Because dicentric chromosomes are known to be unstable [49] , this may be why we could not detect such type of chromosomal rearrangement . We noted that BIR initiation at HMR by the broken chromosome VII would restore an HO cleavable site that might not be properly silenced in the resulting translocations . In the latter case , HO cleavage would destabilize these translocations and their stabilization would require repair by gene conversion ( GC ) using the non-cleavable sequences at MATa-inc . Observed chromosomal rearrangements likely occurred by BIR through the invasion of the MATa-inc and HMR loci . Homology of the centromeric 198-bp Ya-Z1 fragment present at one DSB end would be used to invade MATa-inc or HMR loci at the Ya region and to duplicate chromosome III sequences until reaching its right telomere , generating T7/3-MAT and T7/3-HMR NRTs ( type a , Figure 3 ) . Alternatively , since we observed the cleavage of chromosome III as a consequence of HMR invasion , it is also possible that T7/3-HMR translocations occurred by BIR followed by single-strand annealing ( SSA ) , co-segregating with an intact chromosome III ( type b , Figure 3 ) . Finally , cleaved chromosome III could circularize and co-segregate with the translocation ( type c , Figure 3 ) . To confirm genetically the involvement of BIR in the generation of translocations using our experimental system , we assayed mutants of Rad52 , Rad51 and Pol32 as representative key functions in BIR . Next , we asked whether DNA nucleases acting on branched structures such as D-loops , replication forks or HJs would have a role in the cascade of recombination events that led to translocations . We chose to study the genetic role of Mus81 , Rad1 , Yen1 , Slx1 and Slx4 because of their known functions in recombination processes [25] . Because none of the genes coding for these SSEs are essential for viability , we assayed deletion mutants and combined mutants between them to assay redundancy of functions . Although Slx1 and Slx4 form a heterodimer complex , such as Mus81-Mms4 and Rad1–Rad10 , we decided to study slx1Δ and slx4Δ mutations separately because Slx4 seems to have additional roles in DSB repair apart from regulating the Slx1 nuclease activity [50] , [51] . We also included the DNA helicase Sgs1 because of its known role in HJ dissolution ( Figure 4A ) . For each mutant assayed , we recovered survivors after HO endonuclease induction and selected colonies that contained translocations between chromosome VII and chromosome III , as described above ( 5-FOA-resistant colonies ) . We did not recover any translocants in rad52Δ and rad51Δ mutants , in which strand invasion fails to occur ( Figure 4A ) . POL32 deletion reduced 4-fold the frequency of translocations in comparison with the WT ( χ2 , p<0 . 01 ) , arguing in favor of the involvement of the Pol32-dependent BIR pathway in the generation of the translocations analyzed in this study . We did not observe any increase of translocations in sgs1Δ cells , even though SGS1 gene had been identified as a suppressor of translocations involving template switching events [19] . Among the nuclease single mutants , only mus81Δ showed a slight but significant decrease in the frequency of translocants ( χ2 , p<0 . 01 ) when compared to the WT . In contrast , we observed significant decreases in the frequency of translocants between the WT , mus81Δ rad1Δ and mus81Δ rad1Δ slx1Δ mutants , but these effects were found to be epistatic with mus81Δ ( Figure 4A; χ2 , p<0 . 01 ) . On the contrary , we observed further significant decreases ( χ2 , p<0 . 01 ) in mus81Δ yen1Δ mutants , of about 3 . 5-fold and 2 . 5-fold compared to the WT and mus81Δ , respectively ( Figure 4A ) . We conclude that Mus81 and Yen1 are both required for translocations in our assay . Since it existed the possibility that the translocations were also produced by BIR/SSA , we assayed translocation formation in an HMR-inc strain , in which HMR locus would not be susceptible to HO cleavage and translocations would only be produced by BIR . We observed that survival dropped from 82% in the original HMR WT strain to 52% in the HMR-inc WT strain , although the frequency of translocants remained around 40% ( Figure 4B and Figure S1 ) . In the control HMR-inc background , POL32 deletion reduced 18 . 5-fold the frequency of translocations in comparison with the WT ( Figure 4B; χ2 , p<0 . 01 ) . This result demonstrates a clear dependency of translocations on the Pol32-dependent BIR pathway in this background . We observed a significant decrease of 3 . 4-fold in HMR-inc mus81Δ yen1Δ mutants compared to the HMR-inc WT but no decrease in the HMR-inc mus81Δ and HMR-inc yen1Δ single mutants ( Figure 3B; χ2 , p<0 . 01 ) . These results indicate that Mus81 and Yen1 have overlapping functions in BIR . To confirm this , we performed time-course experiments to monitor the kinetics of appearance of BIR intermediates in these mutants ( Figure 4C ) . Kinetics of DSB formation in the mutants was similar to the WT , allowing us to directly compare the accumulation of BIR intermediates at each time point ( Figure 4D ) . In HMR-inc WT cells , 7/3-HMR and 7/3-HML intermediates appeared 4 h and 5 h after DSB induction , respectively ( Figure 4C , 4D ) . 7/3-MAT intermediates were detected 6 h after DSB induction ( Figure 4C , 4D ) , showing that their delayed appearance in the original HMR WT strain was partly due to HMR cleavage . No BIR intermediate could be detected in HMR-inc pol32Δ cells ( Figure 4C , 4D ) . In HMR-inc mus81Δ yen1Δ cells , a decrease of BIR intermediates was reproducibly observed at all time points ( Figure 4C , 4D ) . We concluded that Mus81 and Yen1 are both required for promoting efficient Pol32-dependent BIR . We confirmed by PFGE that all 5-FOA-resistant survivors analyzed in mutant backgrounds contained chromosome translocations between chromosome VII and chromosome III ( Figure 5A and Figures S5 , S6 ) . Notably , a very low amount of T7/3-MAT translocations ( 8% ) were recovered in WT cells ( Figure 5B ) . In pol32Δ cells , no T7/3-MAT translocation was recovered ( n = 59 , χ2 , p<0 . 05 ) , showing that this type of translocation has a complete dependency on Pol32 ( Figure 5A , 5B ) . Among the nuclease mutants , rad1Δ showed a significant increase of T7/3-MAT translocations , up to 20% of total translocations ( 2 . 5-fold , n = 59 , χ2 , p<0 . 01 ) , which was not observed in double mutants with mus81Δ and slx1Δ ( Figure 5A , 5B ) . The concomitant absence of Mus81 , Slx4 and Yen1 engendered an even higher increase ( 3 . 1-fold ) of T7/3-MAT translocations that reached up to 25% of total translocations ( n = 58 , χ2 , p<0 . 01 ) ( Figure 5A , 5B ) . In kinetic experiments , both 7/3-HMR and 7/3-HML intermediates accumulated in the rad1Δ and mus81Δ slx4Δ yen1Δ mutants with kinetics clearly delayed ( 2–3 h ) respect to the WT , demonstrating a defect of repair in these mutants ( Figure 6A , 6B ) . At 24 h , signals for 7/3-HMR , 7/3-HML and 7/3-MAT that likely correspond to final repair products were detected in all strains . Notably , we observed a clear increase in the 7/3-MAT/7/3-MAT ratio , up to 24% and 29% in rad1Δ and mus81Δ slx4Δ yen1Δ , respectively , compared to the 15% seen in the WT ( Figure 6C ) . This observation correlated with the increase of T7/3-MAT translocations previously observed in these mutants ( Figure 5B ) . The defects observed in rad1Δ and mus81Δ slx4Δ yen1Δ mutants could be explained by the known functions of Rad1 and Slx4 in SSA [50] , which would be required for T7/3-HMR translocation formation by BIR/SSA . Indeed , in the control HMR-inc WT strain , the T7/3-MAT translocations represented about 55% of the translocations observed in 5-FOA-resistant survivors ( Figure S7B ) . These results confirmed that the cleavage of the HMR locus upon its invasion in the HMR WT strain facilitated the formation of T7/3-HMR translocations as opposed to T7/3-MAT translocations . Additionally , possible cleavage of HMR upon the passage of the BIR fork initiated at MATa-inc impaired the formation of T7/3-MAT translocations . Nevertheless , slx4Δ single mutants did not show any increase of T7/3-MAT translocations and mus81Δ slx4Δ yen1Δ mutants showed a higher increase of T7/3-MAT translocations than rad1Δ mutants ( Figure 5B ) . We hypothesized that this was due to a processing defect of BIR intermediates formed at MATa-inc that impeded subsequent template switching to HMR and HML . To explore this possibility , we analyzed the kinetics of appearance of BIR intermediates in mus81Δ slx4Δ yen1Δ mutants , as well as in the rad1Δ and slx4Δ single mutants , in the HMR-inc background ( Figure 6D , 6E ) . In HMR-inc strains , BIR/SSA does not occur and the accumulation of 7/3-HML intermediates serves as an indicator of the efficiency of template switching during BIR . In contrast to the HMR-inc rad1Δ and HMR-inc slx4Δ mutants , which did not show any significant difference compared to the HMR-inc WT strain , 7/3-HML intermediates were reproducibly not detected at all time points in HMR-inc mus81Δ slx4Δ yen1Δ mutants . Accumulation of 7/3-HMR and 7/3-MAT intermediates was also significantly lower in the latter strain , probably reflecting the BIR defect previously observed in HMR-inc mus81Δ yen1Δ cells . Altogether , these data indicate that DSB repair was altered in cells lacking Rad1 , defective in BIR/SSA , and in cells lacking all three SSE factors Mus81 , Slx4 and Yen1 that show a defect of template switching during BIR . Among the chromosomal rearrangements generated in our strains , circularization of chromosome III occurred at a high frequency . No chromosome III circularization was observed in the control HMR-inc WT strain ( Figure S7B ) , demonstrating that HMR cleavage induced this secondary recombination event ( type c , Figure 3 ) . HMR cleavage and translocations via BIR/SSA would leave chromosome III with a one-ended DSB . Thus , circularization of chromosome III is thought to occur by a recombination event that ended in a half-crossover , characterized by a reciprocal exchange between HMR and HML that caused the loss of the chromosome III left telomere and the formation of a chromosome circle ( Figure 7A ) . Circularization of chromosome III is not mandatory for survival of BIR/SSA-mediated translocants since the translocation can co-segregate with the uninvolved chromatid of chromosome III ( type b , Figure 3 ) . Therefore , we took advantage of chromosome III circularization events to investigate the ability of SSE mutant cells to produce half-crossovers . PFGE analyses allowed us to detect circular chromosome III in all strains as signals appearing in the wells ( Figure 7B and Figures S8 , S9 ) . In addition to the well signal , we also detected in some strains a faint signal for the truncated linear chromosome III . This faint signal corresponds in size to the circular chromosome III cleaved by AsiSI ( Figure 1E ) . Since we did not find any linkage between the presence of this signal and a particular genetic background , we conclude that this is likely due to breakage occurring during DNA extraction or PFGE . To get further insight into the mechanism that gave rise to circular chromosomes , we have evaluated the percent of translocants that contained circular versus linear chromosomes III for each genotype ( Figure 7C ) . Whereas about 50% of WT translocants contained a circular chromosome III , this percent increased significantly to 88% ( n = 59 , χ2 , p<0 . 01 ) in pol32Δ mutants ( Figure 7B , 7C ) . This result is concordant with previous observations showing that pol32Δ defects led to D-loop processing during BIR that generated half-crossovers [22] , [23] . Among the SSE mutants tested , the proportion of translocants containing circular chromosomes III decreased significantly to 33% and 30% in mus81Δ and slx1Δ mutants , respectively ( χ2 , p<0 . 05 ) ( Figure 7B , 7C ) . In contrast to this observation , the percentage of translocants containing circular chromosomes III increased significantly up to 72% in mus81Δ slx4Δ double mutants ( χ2 , p<0 . 01 ) ( Figure 7B , 7C ) . These results suggest that Mus81 and Slx4 have different roles regarding crossover formation . Additionally , we observed that yen1Δ mutation suppressed the increase detected in mus81 slx4Δ mutants , suggesting that Yen1 action may be possible only in the absence of Slx4 . We conclude that Slx4 , which has been described as acting as a platform with different nuclease complexes [52] , might regulate Mus81 and Yen1 accessibility to recombination intermediates or their nuclease activity to generate crossovers . We also analyzed chromosome III circularization in Ura+ survivors , which did not contain translocations and likely performed DSB repair by GC or NHEJ ( Figure S10 ) . Analogously to what we observed in the Ura− translocants , 36% of WT Ura+ survivors contained a circular chromosome III and this percentage increased significantly to 90% in pol32Δ mutants ( Figure S10E ) . This suggests that chromosome III circularization does not depend on the translocation event and happens similarly in all survivors whether Ura+ or Ura− . Finally , we calculated the efficiency of chromosome III circularization for each genotype ( Figure 7D ) . Pol32 appears to be required for efficient chromosome III circularization since the calculated efficiency went down significantly from 22% in WT to 10% in pol32Δ translocants ( χ2 , p<0 . 05 ) ( Figure 7D ) . This probably explains the low survival of pol32Δ mutants to the DSB induction ( Figure S1 ) . Similarly , we observed a significant decrease in the efficiency of chromosome III circularization in mus81Δ yen1Δ mutants ( Figure 7D ) . This is consistent with the conclusion that Mus81 and Yen1 play an important role in the formation of the majority of circular chromosomes III .
An important aspect that remains unclear is the apparent instability of the BIR fork , which may be cleaved during its advance , promoting template switching or producing half-crossovers . All essential DNA replication factors except those for pre-replication complex assembly are required for BIR [14] , playing in favor of fork stability . However , the DNA damage checkpoint is activated during BIR and dNTP levels are elevated to facilitate repair , which is thought to happen in the G2 phase of the cell cycle [11] , [13] , [55] . Consequently , DNA synthesis during BIR has been found to be highly inaccurate [56] and replication fork progression may be perturbed by the absence of S phase-specific factors . Interestingly , the nonessential DNA polymerase δ subunit Pol32 seems to represent a key factor for BIR completion but performs a function that is still unknown . In our assay , we observed that pol32Δ mutants had a clear defect in producing translocations , but not as strong as the one observed in rad52Δ and rad51Δ mutants ( Figure 4A ) , confirming that some translocations were produced via BIR/SSA , in which extensive DNA synthesis would not be required . However , Pol32 became essential for translocations in the absence of HMR cleavage in the control HMR-inc strains ( Figure 4B ) . In these strains , we could not detect any BIR intermediate ( Figure 4C ) , meaning that Pol32 is required for DNA synthesis of few kilobases and that the latter was necessary for template switching . Surprisingly , we observed in pol32Δ survivors an extremely high level of chromosome III circularization , even in Ura+ survivors that likely repaired the DSB on chromosome VII by GC or NHEJ ( Figure 7 and Figure S10 ) . This would mean that DSBs induced at HMR could not be repaired by HR with MATa-inc sequences in pol32Δ cells . Indeed , preferential formation of crossovers between MAT and HMR would lead to the extrusion of genes essential for viability , whereas crossovers between HMR and HML would lead to the formation of stable circular chromosomes . Preferential processing of recombination intermediates into crossovers in pol32Δ mutants have been reported in other studies , in which BIR events were aborted and resulted in half-crossovers [22] , [23] . We determined the role of SSEs in the generation of chromosomal rearrangements using our assay as the goal to identify the nucleases that are required during BIR . Overall , we observed a significant decrease of chromosomal rearrangements in mus81Δ single mutants that was aggravated in mus81Δ yen1Δ mutants ( Figure 4A ) . We have confirmed that the frequency of translocants only decreased in the mus81Δ yen1Δ mutants in the HMR-inc background ( Figure 4B ) , in which BIR , and not BIR/SSA , is expected to occur . This indicates that Mus81 may play a role in BIR/SSA and that Mus81 functions can be fully taken over by other proteins during BIR . However , our data show that both Mus81 and Yen1 carry out redundant or equivalent activities , which are needed for BIR completion . Mus81 and Yen1 have already been implicated in DSB repair by recombination but not directly in BIR . Mus81 has been shown to act at replication forks . It has been proposed that Mus81 could cleave stalled forks but also to participate in recombination-mediated repair of cleaved or collapsed forks to allow their restart in yeast and humans [32]–[34] , [57] . Mus81 is also required in humans for telomere recombination to allow proliferation of telomerase-negative cancer cells [58] . Formally , both mechanisms of replication fork restart and telomere recombination are equivalent to BIR . Yen1 roles in recombination have been revealed in the absence of Mus81 . While yen1Δ mutants are repair proficient , mus81Δ yen1Δ double mutants exhibit a higher sensitivity to DNA-damaging agents that disturb replication fork progression than mus81Δ mutants [28] , [30] , [31] . Together , these data point out that Mus81 and Yen1 may promote a replication fork restart mechanism . In vitro , Yen1 is a specialized Holliday junction resolvase [26] whereas the Mus81-Mms4 complex prefers branched DNA substrates that contain a discontinuity or a nick adjacent to the branch point , but also cleaves normal HJs [27] , [59]–[61] . Here , we have demonstrated genetically that both Mus81 and Yen1 were required for efficient BIR . According to previously published data , we propose that Mus81 and Yen1 would act to establish the replication fork required for BIR by processing recombination intermediates such as D-loops or HJs . Nevertheless , BIR still occurred at a low frequency in mus81Δ yen1Δ mutants , suggesting that other factors could promote this critical step of BIR in the absence of Mus81 and Yen1 . Our results are consistent with additional roles of Mus81 and Yen1 in later steps of BIR . We have demonstrated that Mus81 , Slx4 and Yen1 were required together for efficient template switching during BIR . The mus81Δ slx4Δ yen1Δ mutants showed an increased occurrence of T7/3-MAT translocations ( Figure 5B ) , which we infer as being partly due to a defect in template switching from MAT to the HM loci ( Figure 6 ) . However , we did not observe any increase of T7/3-MAT in mus81Δ slx1Δ yen1Δ mutants , even though Slx1 is the catalytic subunit of Slx1-Slx4 nuclease heterodimer . On the contrary , we observed a WT or decreased level of T7/3-MAT translocations in all slx1Δ mutants . We concluded that Slx4 and Slx1 act independently in BIR , presumably because of Slx4 additional functions apart from Slx1 at the replication fork [51] , [62] , [63] . Regarding the involvement of SSEs in half-crossover production , the absence of Mus81 or Slx1 significantly decreased the amount of circular chromosomes III among translocants ( Figure 7C ) . Notably , no further decrease was observed when removing Mus81 , Slx1/Slx4 or Yen1 , all of which have been involved in crossover formation during meiosis in yeast [64] , [65] . This could be due to the involvement of other nucleases such as Mlh3 and Exo1 , as recently reported during the revision of this manuscript [65] . Our assay does not permit a direct analysis of the role of SSEs in half-crossover since the formation of circular chromosomes III is limited by the frequencies of translocations , template switching and HMR cleavage . In principle , HML could also be cleaved upon invasion so that an HMR/HML double cleavage could lead to a circular chromosome III by an SSA-like mechanism . However , this hypothesis is not supported by our results as we observed a similar frequency of circular chromosomes III in rad1Δ mutants and WT . Instead , circularization of chromosome III via SSA would generate a heterologous single-stranded DNA overhang that would require Rad1 for its removal [66] and , indeed , we have observed a requirement of Rad1 in the formation of T7/3-HMR translocations via BIR/SSA in our assay ( Figure 5 and Figure 6 ) . Despite the limitations of our assay , our genetic data suggest an interesting interaction between Mus81 , Yen1 and Slx4 SSEs . Whereas mus81Δ translocants showed a decrease in the frequency of circular chromosomes III , mus81Δ slx4Δ translocants showed a significant increase , which was suppressed in the additional absence of Yen1 ( Figure 7C ) . These results suggest that Slx4 may have a specific role in regulating the ability of Mus81 and Yen1 to catalyze half-crossovers . It has been previously shown that Mus81 is involved in half-crossovers following BIR [23] and that Mus81 and Yen1 independently promote crossovers during gene conversion , Yen1 serving as a backup function in mus81Δ cells [30] . However , here we uncover two parallel pathways , one using Mus81 and Slx4 and the other Yen1 . This is in agreement with a similar involvement recently described for these nucleases in two pathways of crossover formation during sister-chromatid recombination [67] . It remains unclear how Slx4 may regulate Mus81 and Yen1 . A recent cell-cycle analysis of Mus81-Mms4 and Yen1 revealed that their catalytic activities are regulated by phosphorylation events . In mitotic cells , Mus81-Mms4 is hyperactivated by Cdc5-mediated phosphorylation at G2/M while Yen1 is activated later by dephosphorylation in M phase [27] . Nevertheless , it remains unknown if Yen1 can be activated earlier in the absence of Mus81 or upon DSB induction . Mec1/Tel1 kinases phosphorylate Slx4 in response to DNA damage [50] , [51] and may participate in modulating context-specific protein interactions between Slx4 , Mus81 and Yen1 and allow substrate accessibility to activated Mus81 and Yen1 . Altogether , our results suggest that Slx4 plays a central role during BIR . Slx4 may regulate Mus81 and Yen1 , whose cleavage activities are required for replication fork establishment and could either cause template switching or half-crossovers . In the case of one-ended DSBs , it has been proposed that dynamic displacement of the invading strand out of the D-loop would contribute to template switching [16] . This implies that the invading strand would be displaced early during BIR , after a short tract of DNA synthesis . Nevertheless , events of template switching have been observed in later steps of BIR , as far as 10-kb downstream of the site of invasion [16] , [23] . Such a synthesis would expose long tracts of single-stranded DNA if it were the result of the sole extension of the invading strand . Despite the fact that such long single-stranded DNA tails have been involved in gene conversion events monitored in mitotic gap repair assays , we propose that , at some point , priming of lagging strand synthesis would ensure a better protection of the recombination intermediates , safeguarding genome stability . Thus , we propose that template switching events would happen after the establishment of the BIR fork and priming of lagging strand synthesis . In vitro data showed that canonical replication forks are among the preferred substrates of Mus81-Mms4 and Yen1 [25] , [68] , therefore we propose that Mus81 , Slx4 and Yen1 would act on the replication fork during BIR to cause template switching and half-crossovers . Our results together with previous data permit us to propose a new model for BIR and the role of the different SSEs used in this study ( Figure 8 ) . During HR , priming of synthesis from the 3′ invading end extends the initial D-loop and failure to capture the other DSB end would promote BIR . We propose that Mus81 would cleave the extended D-loop structure to allow the establishment of a replication fork . In the absence of Mus81 , branch migration of the D-loop would create an intact Holliday junction , which could be processed by Yen1 with the same outcome . Pol32 would promote extensive DNA synthesis and complete replication would generate a non-reciprocal translocation ( NRT ) . We propose that the BIR fork could stall and be processed by Mus81-Slx4-Yen1 to cause template switching ( [a] , Figure 8 ) . Differential cleavage of the BIR fork by Mus81-Slx4-Yen1 would terminate BIR at the expense of a half-crossover ( [b] , Figure 8 ) . Altogether , this work brings a clearer view about the involvement of SSEs in the BIR mechanism of DSB repair . Importantly , we show that SSEs are involved in replication template switching and half-crossovers , which generate complex chromosomal rearrangements and prolonged cycles of genomic instability . Such events are thought to be at the origin of various genomic disorders and cancer development [24] , [69] .
All Saccharomyces cerevisiae yeast strains used in this study are in W303-1a background ( his3-11 , 15 leu2-3 , 112 , trp1-1 ura3-1 ade2-1 can1-100 rad5-535 ) [70] and harbor MATa-inc , ade3Δ::gal-HO and leu2Δ::SFA1 alleles [53] . The MATa-inc HMR-inc strain was obtained by mating-type switching inducing HO expression in a MATα HMR-inc strain . Independent survivors to HO expression were selected and the MAT and HMR loci were sequenced to verify the MATa-inc HMR-inc genotype . Deletion mutants were either obtained by the PCR-based gene replacement method ( verified by PCR and Southern ) or by genetic crosses ( verified by tetrad analysis ) . Deletion of MAT is only partial ( matΔYZ ) because of the presence of other genes overlapping with MAT . Only the Y and Z sequences , containing the HO cut site , have been removed . Insertion of a HO-cleavable 240-bp HMR fragment at the ADH4 locus has been conducted as follows . Two 5′ and 3′ADH4 fragments were amplified by PCR with the following primer pairs ADH4-5′#1 GCGCGCGGTACCGAATTCAAACCGCTGATTACATCAAA and ADH4-5′#2 GCGCGCAGATCTATCGATCTCGAGTCTAGACTAGACCAGTAGCAGCAGTC , and ADH4-3′#1 GCGCGCAGATCTGCTAGCACTAGTGGATCCCTTAGTCGCTGCATACAAAG and ADH4-3′#2 GCGCGCGAGCTCGAATTCGCACACGCATAATTGACGTT . These two fragments were cloned by the gap repair method in pBluescript II ( SK+ ) previously digested by KpnI and SacI to create pBP99 plasmid . A BglII-BamHI URA3 containing fragment from sp392 plasmid [71] was then cloned in pBP99 digested with BglII to create pBP102 . Finally , the HMR fragment was amplified from genomic DNA with the primer pair HO-HMR-Hind3 GCGCGCAAGCTTCAACCACTCTACAAAACCAAAACCA and HO-HMR-Nhe1 GCGCGCGCTAGCAGAAGAAGTTGCAAAGAAATGTGGC and cloned into pBP102 after digestion with HindIII and NheI , to create pBP102-HO . pBP102-HO was linearized with PvuII and transformed into yeast . Integration was selected by uracile prototrophy and verified by PCR and Southern analysis . Yeast cells were grown in yeast extract-peptone-adenine-dextrose ( YPAD ) until reaching the exponential phase of growth , appropriately diluted with H2O and plated on synthetic complete ( SC ) medium containing either 2% glucose or 2% galactose as a carbon source . Survivor colonies on galactose-containing plates were then restreaked on YPAD plates and replica-plated on SC plates containing 5-FOA ( USBiological ) , a drug that generates a toxic metabolite in Ura+ cells , or on SC plates lacking uracil . Frequencies were calculated as follows: survival frequency = cfu galactose/cfu glucose; translocants frequency = ( cfu galactose Ura−5-FOAr+ ( cfu galactose Ura+5-FOAr ) /2 ) /cfu glucose . 96 to 288 survivor colonies , recovered from 2 to 3 independent induction experiments , were analyzed for each strain tested . Statistical analysis was performed using the χ2 test with Yates' correction . Yeast cells were grown at 30°C in liquid YPAD until reaching the exponential phase of growth , washed twice with synthetic complete medium SGL ( 3% glycerol , 2% lactate ) and cultured overnight in SGL until reaching an OD600 nm≈0 . 5 when galactose was added at a final concentration of 2% . Cells were taken at different times after galactose induction and genomic DNA was extracted in agarose plugs according to standard procedures . Agarose plugs were incubated twice in 200 µl 1× β-Agarase I reaction buffer for 30 min , melted at 65°C for 10 min , equilibrated at 42°C for 15 min and treated with β-Agarase I ( New England BioLabs ) at 42°C for 1 h before PCR amplification . These were performed with 250 ng of genomic DNA ( estimated with NanoDrop , Thermo Scientific ) in a total volume of 30 µl in the following conditions: 1× Phusion HF buffer , 200 µM each dNTP , 0 . 6 U Phusion DNA polymerase ( Finnzymes ) , 0 . 5 µM each primer . Samples were denatured for 45 s at 98°C , then cycled 25–35 times with 20 s denaturation ( 98°C ) , 30 s annealing ( 57°C ) and 45 s extension ( 72°C ) followed by a final extension step of 5 min at 72°C . PCR was performed with primer p7 GCACACGCATAATTGACGTT and primers p3-M GAAGACTTGTGGCGAAGA , p3-R CCAACATTTAGGAAAAAACG or p3-L CGGATGGCACAAGGAACACGCATTT . Control PCR was performed with primers corresponding to ACT1 locus , ACT1up TTCACGCTTACTGCTTTTTTC and ACT1low CAAGGCGACGTAACATAGTTT . PCR products were subjected to gel electrophoresis in 0 . 8% agarose and stained with ethidium bromide . Instead of β-Agarase I treatment , plugs were digested with 30 U of EcoRV restriction enzyme for 5 h at 37°C and loaded in a 1% agarose gel for Southern analysis . Electrophoresis was run at 80 V for 16 h30 and DNA was transferred into Hybond-XL membranes ( GE Healthcare ) in alkaline conditions . Membranes were probed with dCT32P-labelled PCR fragments obtained with ADH4-3′#1 and ADH4-3′#1 primers ( 7L probe ) . Quantification of DNA signals was made relative to the total DNA of each lane and was performed using ImageGauge 4 . 2 ( Fujifilm ) program . For each strain , 28 to 84 independent Ura- 5-FOAr survivor colonies were grown in 2 , 5 ml of YPAD medium overnight at 30°C . Agarose plugs containing chromosomal DNA were made according to the manufacturer's instructions ( Bio-Rad ) . AsiSI digestion was performed incubating agarose plugs twice in 1 ml 1× NEBuffer 4 for 30 min and digested in 200 µl 1× NEBuffer 4 with 30 U of AsiSI restriction enzyme for 5 h ( New England Biolabs ) . Agarose gels ( 0 . 9% ) were run in a Bio-Rad CHEF MapperXA apparatus for 16 h at 6 V/cm with a switch time of 70 s and for an additional 12 h at 6 V/cm with a switch time of 120 s . Then , gels were stained with ethidium bromide and DNA was transferred into Hybond-XL membranes ( GE Healthcare ) in alkaline conditions . Membranes were probed with dCT32P-labelled PCR fragments obtained with primers ADH7#1 TGTTGGCTAAAGCTATGG and ADH7#2 TTCTTCGCTGATCGG ( 3R1 probe ) , ARS315#1 AAACCAGTCTTTAACCGCCATAATG and ARS315#2 CAGAGCCCAAGAGATAGCCGAACTT ( 3R2 probe ) , and with primers HML+HMR-F CAAACATCTTAGTAGTGTCTGAGGA and HML+HMR-R CTGTAATTTACCTAAGTTACCAGAG ( X probe ) . Chromosomal rearrangements different from T7/3-MAT or T7/3-HMR translocations or circular chromosomes III and revealed by the PFGE analysis were not included in the statistical analyses . | Genome rearrangements consisting of non-reciprocal translocations ( NRTs ) seem to play an important role in carcinogenesis in humans . They are likely caused by intracellular mechanisms that are normally committed to repair breaks occurring in the DNA molecule . Failure of faithful repair of DNA double-strand breaks ( DSBs ) often leads to chromosomal rearrangements when repair occurs within repeated genomic regions . The break-induced replication ( BIR ) pathway of DSB repair is a major source of complex chromosomal rearrangements , the latter occurring when BIR involves template switching between dispersed repeated sequences . Given the deleterious consequences of such events for genomic stability , it is of great significance to understand the molecular bases of BIR . Here , we examined the role of different DNA nucleases in chromosomal rearrangements and uncovered the functional involvement of the structure-selective endonucleases ( SSEs ) subunits Mus81 , Yen1 , and Slx4 at different steps during BIR . Our work provides new clues to understand the origin of NRTs and the role of SSEs in their generation . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"genetics",
"biology",
"genomics",
"molecular",
"cell",
"biology",
"genetics",
"and",
"genomics"
] | 2012 | Complex Chromosomal Rearrangements Mediated by Break-Induced Replication Involve Structure-Selective Endonucleases |
Understanding of genotype-phenotype associations is important not only for furthering our knowledge on internal cellular processes , but also essential for providing the foundation necessary for genetic engineering of microorganisms for industrial use ( e . g . , production of bioenergy or biofuels ) . However , genotype-phenotype associations alone do not provide enough information to alter an organism's genome to either suppress or exhibit a phenotype . It is important to look at the phenotype-related genes in the context of the genome-scale network to understand how the genes interact with other genes in the organism . Identification of metabolic subsystems involved in the expression of the phenotype is one way of placing the phenotype-related genes in the context of the entire network . A metabolic system refers to a metabolic network subgraph; nodes are compounds and edges labels are the enzymes that catalyze the reaction . The metabolic subsystem could be part of a single metabolic pathway or span parts of multiple pathways . Arguably , comparative genome-scale metabolic network analysis is a promising strategy to identify these phenotype-related metabolic subsystems . Network Instance-Based Biased Subgraph Search ( NIBBS ) is a graph-theoretic method for genome-scale metabolic network comparative analysis that can identify metabolic systems that are statistically biased toward phenotype-expressing organismal networks . We set up experiments with target phenotypes like hydrogen production , TCA expression , and acid-tolerance . We show via extensive literature search that some of the resulting metabolic subsystems are indeed phenotype-related and formulate hypotheses for other systems in terms of their role in phenotype expression . NIBBS is also orders of magnitude faster than MULE , one of the most efficient maximal frequent subgraph mining algorithms that could be adjusted for this problem . Also , the set of phenotype-biased metabolic systems output by NIBBS comes very close to the set of phenotype-biased subgraphs output by an exact maximally-biased subgraph enumeration algorithm ( MBS-Enum ) . The code ( NIBBS and the module to visualize the identified subsystems ) is available at http://freescience . org/cs/NIBBS .
Certain industrial processes , such as the production of hydrogen and ethanol , benefit from using prokaryotic or eukaryotic organisms to produce , reduce , and convert important chemical compounds [1] , [2] . Bioengineers search for ways to modify phenotypic traits , or phenotypes , of these organisms to improve the overall process efficiency [3] . Modifications to the organism's phenotype are made through modifications to its genome . In order to obtain the desired changes in the organism's phenotype , engineers require a deciphering of which genes are related to the expression of the given phenotype , also known as genotype-phenotype associations [4] , [5] . Unfortunately , such an understanding has not kept pace with the rate at which genes are discovered [6] . Uncovering genotype-phenotype associations could be greatly improved if organism's metabolic systems involved in the phenotype expression were understood [7] . These systems involve multiple metabolic reactions that are grouped into functionally-distinct modules called metabolic pathways [8] . Changes to the enzymes in these modules can affect the expression of the phenotype of interest . Thus , it is imperative to be able to identify all of the enzymes that make up a phenotype-related metabolic system . The task of identifying a phenotype-related metabolic system consists of two main subtasks: determining the metabolic system and establishing that it is phenotype-related . Understanding how a system has been evolutionarily conserved has been used as an approach to accomplish both tasks . If a set of interacting metabolic reactions are important for expressing the target phenotype , then there likely exists an evolutionary pressure to conserve the set as a whole , or to have them co-present together , in multiple organisms [9] . The assumed reason for this evolutionary pressure is that the set forms a metabolic system whose function is required by the organism and by its descendants [9] . This is the motivation behind network alignment and phylogenetic profiling approaches proposed to-date . The former [9]–[13] look for subgraphs that exist in metabolic networks of multiple organisms . The latter [4] , [5] , [14] seek to find genes or enzymes that are more likely to be present in phenotype-expressing organisms than in phenotype-non-expressing organisms due to an evolutionary pressure to conserve the phenotype-related enzymes [14] . However , neither network alignment nor phylogenetic profiling approaches can alone identify phenotype-related metabolic systems . Network alignment algorithms can identify metabolic systems present in all or most of a given set of organisms; such a set is typically small , e . g . , less than 10 networks . However , even if the set of organisms exhibit a common phenotype , current network alignment approaches cannot distinguish phenotype-related metabolic systems from other common metabolic systems . Additionally , network alignment approaches would likely not identify a metabolic system if it is only common to a subset of the organisms being compared . Phylogenetic profiling approaches can identify phenotype-related enzymes that are specific to phenotype-expressing organisms . However , it is possible that enzymes that are part of a phenotype-related metabolic system will not be specific to phenotype-expressing organisms; therefore , these approaches will likely miss them . Additionally , it would be computationally intractable to compare the presence of every possible set of enzymes to the presence of the phenotype . In order to address these and other limitations of existing methods , in this paper , we introduce the Network Instance-Based Biased Subgraph Search ( NIBBS-Search ) algorithm ( Figure 1 ) that enables in silico , fast , and accurate prediction of phenotype-related metabolic systems . The predictions arise from comparative analysis of multiple genome-scale metabolic networks . The approach is capable of predicting phenotype-related metabolic systems that are unlikely to be found by current in silico methods . These include but are not limited to metabolic systems that are specific to a subset of the phenotype-expressing organisms that may exhibit a sub-phenotype of the target phenotype ( e . g . , dark fermentative , light fermentative or bio-photolytic sub-phenotypes of biohydrogen production phenotype ) . A network structure–a maximally-&-phenotypically-biased subgraph ( MPBS ) –is introduced to model phenotype-related metabolic systems in a set of metabolic networks derived for dozens or even hundreds of organisms . To assess NIBBS-Search 's accuracy , we first present the Maximally-Biased Subgraph Enumeration ( MBS-Enum ) method that exactly enumerates all MPBS s in a given set of networks; these subgraphs are then used for comparison with the NIBBS-Search results . To overcome MBS-Enum's computational complexity , NIBBS-Search heuristically approximates the set of MPBS s in the set of networks . NIBBS-Search runs orders of magnitude faster than MBS-Enum , while identifying with high sensitivity subgraphs that are statistically significant approximations of the set of MPBS s . Also , the NIBBS-Search -predicted systems contain known phenotype-related enzymes and pathways , including those that only exist in a subset of the phenotype-expressing organisms .
The NIBBS algorithm identifies phenotypically-biased edges from a metabolic map called the seeds and then expands each seed into a maximally , phenotypically-biased metabolic system . The method requires a set of organisms that express the phenotype of interest and ones that do not . A phenotype-profile vector is built for the organism set ( see Figure 1 ) . This organism phylogenetic profile vector and the organism-specific metabolic maps from the KEGG database [15]–[17] are provided as input to NIBBS . The organism-specific metabolic map is a graph , each edge corresponding to a metabolic reaction , substrates and products as its vertices at the two ends of the edge , and the edge label is the enzyme that catalyses the reaction . NIBBS as its first step identifies the phenotypically-biased edges called seeds . Informally , an edge is phenotypically-biased if it is present in a larger number of phenotype expressing organisms when compared to phenotype non-expressing organisms . The seed edges are then expanded into maximally , phenotypically-biased metabolic subsystems by the addition of other edges from the genome-scale metabolic map . The details are discussed in the Methods section . We identified both phenotype-expressing and phenotype non-expressing organisms via literature search . We primarily analyzed six main phenotypes , aerobic respiration , anaerobic respiration , TCA ( citrate cycle ) expression , rTCA ( reverse TCA ) expression , hydrogen production , and acid-tolerance . We also looked at three sub-phenotypes of hydrogen production: dark fermentation , light fermentation , and bio-photolysis . The summary of the organisms used for each experiment is listed in Table 1 . We used the metabolic networks and enzyme lists available in the KEGG database [15]–[17] . The results for all the experiments are available as supplemental files in the website mentioned in the abstract . In order to predict enzymes related to a microorganism's ability to tolerate low pH conditions , ten acid-tolerant organisms and eight alkaliphiles were analyzed using the NIBBS algorithm ( Table S13 ) . Analysis of the NIBBS enzymes shows that 73% acid-tolerant enzymes were recalled , when acid-tolerant organisms were used as positive instance . NIBBS enzymes predicted 164 enzymes , while the Student's T-Test identified only 17 as phenotype-related . Enzymes identified by the Student's T-Test and missed by NIBBS included enzymes involved in central metabolism , amino acid metabolism , and lactic acid metabolism . Two experiments were performed to measure the ability of the NIBBS algorithm to identify enzymes and potential subpathways related to organisms capable of expressing specific pathways . In order to assess the ability of both approaches to identify phenotype-related enzymes , 36 aerobic organisms and 36 anaerobic organisms were selected . Analysis of the NIBBS enzymes shows 86% and 75% recall , respectively , when one or the other are used as positive instances . The results showed that NIBBS enzymes for aerobic respiration contained 261 enzymes and for anaerobic respiration contained 93 enzymes , while the Student's T-Test identified 131 enzymes for aerobic respiration and 64 enzymes for anaerobic respiration . Examination of the enzymes found by the Student's T-Test but missed by NIBBS-Search shows that they are typically present in most of the phenotype-expressing and non-expressing organisms . The reason some enzymes are identified as phenotype-related by the statistical analysis is due to the fact that they typically have a higher copy number in phenotype-expressing organisms . Since NIBBS-Search uses binary data ( i . e . , whether at least one copy of the enzyme is present in the organism ) , these enzymes are not identified by NIBBS-Search as biased . In addition , because the NIBBS algorithm does not rely on the enzyme distributions across entire sets of organisms , it is capable of identifying subgroups of organisms among the list of given species . As such , it is not expected that NIBBS will contain identical sets of enzymes as those identified with the Student's T-Test approach . Due to the ability of the NIBBS-Search algorithm to predict phenotype-related enzymes through the prediction of phenotype-related metabolic systems , the algorithm is capable of distinguishing between pathways that contain common enzymes . To demonstrate this feature of NIBBS-Search , two experiments were conducted comparing the two well-characterized metabolic networks , tricarboxylic acid ( TCA ) cycle and the reverse TCA ( rTCA ) cycle . Sets of organisms known to utilize the TCA and rTCA cycle were selected and analyzed ( Table S16 ) . Selection of the two metabolic systems was due to the ability of these pathways to utilize the same set of metabolites and have common enzymes . Using sixteen organisms that utilize the TCA cycle and six organisms that utilize the rTCA cycle , NIBBS algorithm was able to identify all but one TCA enzyme , malate dehydrogenase ( EC 1 . 1 . 1 . 37 ) , among the top ranking systems ( Table S17 ) . Malate dehydrogenase is part of another system which also includes seven of the eight TCA enzymes ( isocitrate dehydrogenase is not included ) . All eight of the TCA enzymes are , therefore , part of at least one statistically significant system identified in the TCA experiment . To ensure the sensitivity of the algorithm to identifying key enzymes characteristic for each pathway , we reviewed the results to determine if key rTCA enzymes were present in any of the positive instances . In this study , we did not identify any of the three key enzymes unique to rTCA and this suggests that the NIBBS algorithm was able to properly predict the TCA pathway for phenotype-expressing organisms . Similar results are obtained in the rTCA experiment ( Table S18 ) , when rTCA-utilizing organisms are used as positive instances . A top ranking system identified in the rTCA experiment contains seven of the eight rTCA enzymes , including all the five enzymes that the rTCA cycle shares with the TCA cycle ( Table S16 ) . The rTCA-related enzyme , fumarate reductase ( EC 1 . 3 . 1 . 6 ) was not indicated as present in any system identified in the rTCA experiment . In the rTCA experiment , systems identified by NIBBS include two enzymes , citrate synthase ( EC 2 . 3 . 3 . 1 ) and succinate dehydrogenase ( EC 1 . 3 . 99 . 1 ) that are typically associated with the TCA pathway [58] . This is because these two enzymes are not only present in all of the rTCA expressing organisms in the experiment but also in most , if not all , of the TCA expressing organisms in the experiment . This makes them likely to be included in the set of expansion edges , since they do not decrease the–value of the system . The presence of these TCA-related enzymes in rTCA related systems does not indicate an additional functionality , but rather that succinate dehydrogenase found by KEGG might actually be acting as a fumarate reductase . Being that the two enzymes are evolutionarily related to each other , fumarate reductase and succinate dehydrogenase are difficult to distinguish based on sequence alone [59] . In this method we desribe an experiment that evaluates the accuracy of our method using some specialized metabolic pathway information . For this experiment we chose a group of 13 specialized metabolic pathways ( Text S1 ) to act as an artificial phenotype . We then selected around 130 organims that have all these pathways ( Text S1 ) . We divided the organisms into two groups , one group was called the “P” and the second group was called the “N . ” From the metabolic networks of the organisms belonging to the “N” group , we removed the enzymes that overlap with the chosen metabolic pathways , thus creating an artificial bias . If NIBBS-Search can truely identify phenotype-related subsystems , then it should be able to identify the subsystems related to these metabolic pathways as significant . In fact , we found that all the 13 pathways were significantly present in the discovered subsystems . There are three parameters that the NIBBS algorithm takes as input: ( i ) the percentage of the positive organisms the resulting subsystem ( expanded seed set ) should be be present ( ) , ( 2 ) the maximum bias ( maximum ) , and ( 3 ) the maximum size of the seed set ( ) . All these parameters have been analyzed using the same artificial dataset created using the 13 specialized metabolic pathways discussed in the Systematic Validation section . The paramater is utilized while performing seed-expansion to control in how many phenotype expressing organisms the resulting expanded seed set should be present . is the most stringent value and would require that the resulting subsystem be present in all of the organisms the seed-set was present in . We utilized this value as default to make sure that only the strongest signals are recorded . However , for this experiment we varied the value between and at step intervals to analyze the effect . We found that for smaller values of , the number of subsystems output are fewer when compared to the larger value of . However , for small values the subsystem sizes are larger . This effect is due to the fact that more edges get added during the seed-expansion stage because of the lenient ( small ) threshold . When we looked at the corresponding phenotype-bias values for the identified subsystems , we found that for a , of the systems have phenotype-bias value of less than , this number steadily decreases until where only of the subsystems have significant phenotype-bias . The parameter is the maximum seed set size in a NIBBS run . A would mean that every candidate seed with a less than the maximum becomes its own seed set and then seed expansion is run on each singleton seed set . We utilized the value for our experiments . However , we analyzed the effects of by varying the value between and at step intervals . We foound that except for , NIBBS identified the 13 specialized metabolic pathways to be signficant for all the other values . The maximum bias ( maximum ) value is chosen to provide an upper bound for the bias value of the enumerated subsystems . We varied the maximum bias value between and in step intervals . Fro example , setting the maximum bias value as will enumerate all the subsystems with final bias value of . We found that the number of subsystems produced for a maximum bias value is greater than or equal to the number of subsystems produced for maximum bias value of . The analysis and data related to this section are available in Text S2 . In order to display the dramatic improvement in the runtime of the NIBBS-Search algorithm over exact algorithms , such as MBS-Enum , 98 organism-specific networks are constructed using the global metabolic reference map from the KEGG database [15]–[17] , which contains 1 , 348 vertices and 1 , 476 edges: 50 metabolic networks from aerobic organisms and 48 metabolic networks from anaerobic ones . The MULE algorithm of Koyutürk et al . [63] is used to enumerate maximal frequent subgraphs for all support count thresholds between 1 and the number of positive instances required by MBS-Enum . MULE is selected because both MBS-Enum and NIBBS-Search leverage its network instance model . Such a model allows MULE to enumerate maximal frequent subgraphs by enumerating maximal frequent edge sets , which makes it one of the most efficient methods for enumerating maximal frequent subgraphs [63] . The MBS-Enum is not a wrapper around the MULE algorithm . Even using the efficient MULE algorithm , the runtime of MBS-Enum is intractable for the large-scale networks in this experiment . Figure 5 ( Table S20 ) depicts the MULE runtime for the various thresholds used by MBS-Enum . This runtime grows exponentially , eventually reaching 57 days to enumerate the maximal frequent subgraphs given a support count threshold of 35 . In contrast , the total time required by the NIBBS-Search to approximate the set of maximally-biased subgraphs is 31 seconds ( the dotted line ) . The results in this section describe the typical correspondence between the set of subgraphs output by the NIBBS-Search and the complete set of maximally-biased subgraphs produced by MBS-Enum ( Table S21 ) . To cope with computational intractability of MBS-Enum , only small-size network maps are considered . Specifically , the 33 experiments correspond to the 33 metabolic pathway maps from KEGG that satisfy the two requirements: ( 1 ) all of their maximally-biased subgraphs can be enumerated by MBS-Enum within 24 hours; ( 2 ) a completely random subgraph can be generated by a randomization algorithm at a rate of at least one per second . For each of these 33 network maps , a set of 87 network instances are created . These 87 network instances are divided between 33 positive instances for aerobic organisms and 54 negative instances for anaerobic organisms . Each experiment is labeled with the KEGG pathway identifier ( mapXXXXX ) of the network map used to create the network instances . An approximation score is used to measure the degree to which a set of NIBBS-Search 's subgraphs approximates a set of all maximally-biased subgraphs . The approximation score is calculated by first computing the value for each maximally-biased subgraph in . The value is equal to the maximum Jaccard index ( Equation 1 ) between a maximally-biased subgraph and any subgraph ( Equation 2 ) . The appoximation score is then calculated as the normalized Euclidean distance between the scores computed for the set of NIBBS-Search 's subgraphs and the optimal . ( 1 ) ( 2 ) ( 3 ) Two empirical -values are calculated to determine the statistical significance of the approximation scores . Both -values are calculated as the empirically-determined probability that a set of randomly generated subgraphs would generate a value that is less than or equal to the value of . Each randomly generated set of subgraphs contains the same number of random subgraphs as the set . The random subgraphs used to calculate the -value , , are randomly selected from the set of connected subgraphs in the network map associated with the experiment . For the , the random subgraph of the set is required to be of the same size as the NIBBS-Search's subgraph from the set . By ensuring that the random subgraphs are of the same size as the NIBBS-Search's subgraphs , the calculation of addresses some of the noise that might arise in the -value when the random subgraphs are of a different scale than the NIBBS-Search 's subgraphs . The negative-logs of the empirical values of and are shown for each of the 33 experiments in Figure 6 . As can be seen in Figure 6 , 100 percent of the experiments had a . In addition , 88% of the experiments had a . These results give strong support to the claim that NIBBS-Search identifies subgraphs that are typically close approximations of the set of maximally-biased subgraphs . Thus , if maximally-biased subgraphs are a good model of phenotype-related metabolic systems , NIBBS-Search should be able to identify them as models of phenotype-related metabolic systems .
Application of the NIBBS-Search algorithm to the hydrogen producing and acid-tolerant phenotypes resulted in the prediction of potentially important enzymes , metabolic pathways , and key regulators involved in maintaining or enhancing the production of hydrogen in individual microorganisms . Such predictions include pathways , such as fatty acid biosynthesis , which may help hydrogen producers respond to pH changes both internally and externally . The response to both the formation and uptake of fatty acids present in the surrounding environment suggests that fatty acid biosynthesis could potentially act as a key regulator in metabolic shifts in microorganisms , such as C . acetobuylicum . Other examples provided by NIBBS included the presence or absence of acid tolerant systems and enzymes within specific Clostridium species . In this study , results indicate that C . perfringens contains potentially important enzymes involved in the acid-tolerant ADI pathway . The identified enzymes may then suggest clues necessary for development of gene expression and molecular validation studies . In addition to identifying conserved metabolic pathways , results from the NIBBS algorithm suggest that this method can potentially identify metabolites common to different metabolic pathways . One example of such a metabolite is acetyl-CoA . Acetyl-CoA is generated from pyruvate during glycolysis and can be utilized by differing pathways , including the aerobic TCA cycle and anaerobic formate hydrogen lyase pathway . In the aerobic TCA pathway , the enzyme , pyruvate dehydrogenase , catalyzes the decarboxylation of pyruvate to ( g ) and acetyl-CoA . Acetyl-CoA generated using this process can then be incorporated into the TCA cycle to produce important biosynthetic precursors for other metabolic pathways and energy for microorganisms [34] , [64] . In the anaerobic pathway , pyruvate formate lyase is used to convert pyruvate into acetyl-CoA and formate . Formate produced can then be oxidized by formate hydrogen lyase ( FHL ) to form ( g ) and ( g ) . In the hydrogen studies , the NIBBS algorithm predicts the presence of both pyruvate formate lyase ( E . C . 1 . 1 . 99 . 3 ) and pyruvate dehydrogenase ( E . C . 1 . 2 . 4 . 1 ) when dark fermentative hydrogen producing organisms are compared to hydrogen non-producing organisms . The presence of both pathways may be due to the fact that some dark fermentative microorganisms are capable of utilizing both pathways and the degree to which they utilize each pathway may be dependent on the “cross-talk” between both pathways . However , depending on environmental conditions , the bacteria are grown under , the organism may be more prone to express one phenotype over the other . To understand the role of these pathways , further experimental analysis is required . Identification of common metabolites and potential cross-talk between metabolic pathways is a key step towards understanding metabolic processes , networks , and regulation of phenotype expression in organisms , such as hydrogen producing organisms . While numerous genetic and experimental studies have been conducted to understand the metabolic processes involved in hydrogen production , there is still little understanding of the cross-talk between key hydrogen producing pathways . To help close this gap , biologist could potentially use the NIBBS algorithm to complement hypothesis-driven studies . One way would be to identify phenotype related-pathways , such as the two pathways for acetyl-CoA production , and then conduct molecular studies to review these pathways in organisms shown positive for both pathways . The idea of identifying phenotype-related systems has always been of interest to scientists for many years now and almost all existing methodologies look at phenotypes one at a time . The only method that looks at more than one phenotype , to the best of our knowledge , is the one presented by Liu et al , [65] but even here , the authors primarily look at one phenotype at a time and then use the Pfam-phenotype relationship discovered to identify groups of related phenotypes . Liu et al [65] , however , also do not analyze the effects of multiple phenotypes simultaneously . Clostridium acetobutylicum and Clostridium perfringens have both dark fermenting organisms , but they also share other common phenotypes like anaerobicity and tolerance to acid . These phenotypes , if analyzed as a group , may provide us more information about the phenotype systems in these two organisms than if they were looked at individually . A future improvement could be for NIBBS to analyze multiple phenotypes together . In any comparative genomics , there is always the question whether the identified modules are truely related to the phenotype or they were identified because the organisms are phylogenetically close to each other . Incorporating a method to identify not only phenotypically-biased organisms but also subsyetems present across a phylogenetically diverse group might be one future improvement . This probably can be done by creating a metric that will use the pair-wise phylogetic distances among all the organisms the subsystem is present in . A subsystem present across a phylogenetically diverse group should be scored higher than one that is present across a phylogenetically similar group of organisms . The quality of NIBBS results is also dependent on the underlying data . We discussed one issue in the previous paragraph about phylogenetic diversity . Another issue is the fact that the quality of the results is also dependent on high-quality enzyme-reaction associations . However , databases like KEGG , MetaCyc , and BioCyc provide fairly standard data that can be utilized for such an analysis .
The proposed approach requires a metabolic network model that enables: To satisfy these requirements , we adapt the method of modeling organism-specific networks introduced by Koyutürk et al . [67] . Derived from the KEGG database [15]–[17] , non-organism-specific , yet biochemically feasible , metabolic networks , or reference maps , are modeled as networks whose vertices represent chemical compounds , or metabolites , and whose edges represent reactions that convert metabolites to products . The reaction set corresponds to the set of known reactions that can perform such a conversion . Each reaction is associated with an Enzyme Commission ( EC ) number [68] that is also associated with enzymes that can catalyze the reaction . While metabolic reference maps capture every known , biochemically feasible metabolic process , organism-specific networks describe the metabolic network that exists in a given organism . Specifically , every edge in such a network is associated with an EC number of the enzyme that is known or predicted to be present in the organism . We obtain the organism-specific networks from the reference maps by retaining only those reactions that are catalyzed by an enzyme present in the organism , i . e , by retaining only those edges whose edge labels represent enzymes present in the organism . A subgraph is said to exist in an organism-specific network , if the edge lables , i . e . , the enzymes are present in the organism . Thus , we do not solve any subgraph isomorphism problem . In addition , with this model , the set of all possible metabolic systems can be defined as the set of subgraphs of the reference map . Moreover , only connected subgraphs need to be considered , because metabolic systems are defined as a series of metabolic reactions , where the product metabolites of one reaction are used as the substrate metabolites of the next reaction . The introduced -value of a metabolic system measures the degree of a system's phenotype-bias . It is based on the hypothesis that the systems with the greatest degree of bias ( i . e . , smaller ) will be the systems that are most likely to be phenotype-related . Thus , the search for phenotype-related metabolic systems will aim to minimize the . To calculate the for a given system , the organism-specific networks are divided into two sets: those for phenotype-expressing organisms , or a positive set , and those for phenotype-non-expressing organisms , or a negative set . Given the number of organism-specific networks ( ) , the number of positive networks ( ) , the number of networks that the system exists in ( ) , and the number of positive networks the system exists in ( ) , the phenotype-bias metric is defined according to the cumulative hypergeometic probability distribution: ( 4 ) Because , , , and can be determined given the system subgraph and the set of positive and negative networks , the notation will also be used to describe the phenotype-bias metric . In order to predict phenotype-related metabolic systems , this approach searches the set of organism-specific networks for maximally-biased subgraphs . A maximally-biased subgraph is a subgraph that satisfies the following two criteria: The first criterion comes from the assumption that the entire phenotype-related system is at least as biased as its smaller part . The second criteria is the one that makes the reported subgraphs maximal . According to the second criteria , only allowing those subgraphs that have no larger subgraph with equal or smaller bias are reported . This section presents the Maximally-Biased Subgraph Enumeration ( MBS-Enum ) and the Network Instance Based Biased Subgraph Search ( NIBBS-Search ) algorithms that respectively enumerate the exact and the approximate set of maximally-biased subgraphs as models of phenotype-related metabolic systems . While being exact , MBS-Enum becomes computationally intractable for genome-scale networks . In contrast , NIBBS-Search is a fast heuristic , suitable for hundreds of genome-scale networks; yet , it produces a statistically close approximation of the full set when empirically tested against MBS-Enum results generated for small-scale networks . The hypergeometric test is utilized to identify the pathways enriched by the metabolic subsystems identified by NIBBS for the hydrogen production , dark fermentation , and acid tolerance phenotypes . The enriched pathways are identified for Clostridium acetobutylicum as follows . The edges in all the subsystems are combined into one list and the duplicates are removed . For each metabolic pathway , the edges in the KEGG reference pathway map form the population . The edges in the organism-specific pathway map of become successes in the population . The edges in become the sample and are the successes in the sample . | Genetic engineers often seek to modify the physical traits of microorganisms used in industrial processes in order to improve the efficiency of the overall process . The genes targeted for modification in these cases are typically identified by searching for genes whose presence in an organism is correlated with the presence of the physical trait . In the last few years , however , it has become understood that the physical traits of an organism are often the result of a coordinated set of interactions between multiple genes that make up a biological subsystem . This gives rise to a computational tractability problem , since the number of possible sets of genes is exponentially larger than the number of genes in an organism . Here , we use biological networks to limit the search space to sets of genes known to interact . The presence of the biological subsystems identified by this approach are shown to be significantly correlated to the presence of the phenotype . The results show that this framework can provide potential genetic targets for modifying the expression of a given phenotype . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"algorithms",
"systems",
"biology",
"computer",
"science",
"genomics",
"biology",
"computational",
"biology",
"comparative",
"genomics",
"metabolic",
"networks",
"genetics",
"and",
"genomics"
] | 2012 | NIBBS-Search for Fast and Accurate Prediction of Phenotype-Biased Metabolic Systems |
Whole-genome sequencing technologies are being increasingly applied to Plasmodium falciparum clinical isolates to identify genetic determinants of malaria pathogenesis . However , genome-wide discovery methods , such as haplotype scans for signatures of natural selection , are hindered by missing genotypes in sequence data . Poor correlation between single nucleotide polymorphisms ( SNPs ) in the P . falciparum genome complicates efforts to apply established missing-genotype imputation methods that leverage off patterns of linkage disequilibrium ( LD ) . The accuracy of state-of-the-art , LD-based imputation methods ( IMPUTE , Beagle ) was assessed by measuring allelic r2 for 459 P . falciparum samples from malaria patients in 4 countries: Thailand , Cambodia , Gambia , and Malawi . In restricting our analysis to 86k high-quality SNPs across the populations , we found that the complete-case analysis was restricted to 21k SNPs ( 24 . 5% ) , despite no single SNP having more than 10% missing genotypes . The accuracy of Beagle in filling in missing genotypes was consistently high across all populations ( allelic r2 , 0 . 87-0 . 96 ) , but the performance of IMPUTE was mixed ( allelic r2 , 0 . 34-0 . 99 ) depending on reference haplotypes and population . Positive selection analysis using Beagle-imputed haplotypes identified loci involved in resistance to chloroquine ( crt ) in Thailand , Cambodia , and Gambia , sulfadoxine-pyrimethamine ( dhfr , dhps ) in Cambodia , and artemisinin ( kelch13 ) in Cambodia . Tajima’s D-based analysis identified genes under balancing selection that encode well-characterized vaccine candidates: apical merozoite antigen 1 ( ama1 ) and merozoite surface protein 1 ( msp1 ) . In contrast , the complete-case analysis failed to identify any well-validated drug resistance or candidate vaccine loci , except kelch13 . In a setting of low LD and modest levels of missing genotypes , using Beagle to impute P . falciparum genotypes is a viable strategy for conducting accurate large-scale population genetics and association analyses , and supporting global surveillance for drug resistance markers and candidate vaccine antigens .
Malaria is a major global health burden , with drug resistance in Plasmodium falciparum a major impediment to disease containment and eradication . A deeper understanding of the biology of this parasite and the epidemiology of the disease it causes is needed to inform public health responses , including the timely development of new therapeutics and vaccines [1 , 2] . Interrogating the genetic determinants of P . falciparum virulence has become a crucial aspect of malaria surveillance and control strategies [2 , 3] . A large catalogue of high-density single nucleotide polymorphisms ( SNPs ) in P . falciparum has been identified through whole-genome sequencing studies , enabling population genetics analysis to characterize global parasite diversity [2 , 4] . This advance has also facilitated the discovery of putative genetic determinants of drug resistance phenotypes and candidate vaccine antigens [3 , 5–7] . These and additional SNPs discovered in expanding collections of global P . falciparum clinical isolates [2 , 4] may provide new insights into the diversity and pathogenicity of contemporary parasite populations . The ability to locate naturally-selected regions of this parasite’s genome is critical in uncovering the genetic determinants of drug resistance or candidate vaccines . Several SNPs in P . falciparum genes are known to confer resistance to antimalarial drugs such as chloroquine ( crt [8] ) , sulfadoxine-pyrimethamine ( dhps [9] , dhfr [10] ) , and artemisinin derivatives ( kelch13 [3] ) . Statistically significant signatures of recent positive selection , such as long-range haplotypes and extended-haplotype homozygosity ( EHH ) , have been used to identify genetic mutations contributing to drug resistance in parasite populations . Similarly , the detection of signatures of balancing selection facilitates the identification of parasite targets of acquired immunity . As immunity to the commonest alleles rises in malaria-endemic areas , rarer alleles confer parasites a selective advantage and correspondingly increase in frequency , thereby displacing previously common alleles . This process maintains a balance of alleles in the population , with neither the common alleles moving to fixation nor the rare alleles moving to extinction . Balancing selection is detected using statistics such as Tajima’s D [11] and is evident in vaccine targets , including that encoded by ama1 [12] . P . falciparum has a complex life cycle comprising a haploid phase in humans and a diploid phase in Anopheles mosquitoes , which involves gamete fertilization and sexual recombination [13] . Genetic variants in the P . falciparum genome ( 23 Mb , 14 chromosomes , 81% AT content , ~5500 genes ) were initially surveyed using Sanger sequencing on a limited set of laboratory-adapted clones [14 , 15] . Although recent genomic scans use next-generation sequencing technologies [4 , 6] , high AT content and technological limitations lead to uneven genomic coverage; thus , a proportion of SNP genotypes may be called as missing , which compromises downstream analysis . For example , typical approaches exclude SNPs and parasite isolates with high rates ( >10% ) of missingness from subsequent analysis [4] . Harnessing the full potential of fine-scale population genetics methods , such as haplotype scans for signatures of natural selection , requires complete genotype data on a dense set of SNPs in a large number of samples . The standard approach of restricting analysis only to SNPs without missing data may be problematic . Low levels of individual SNP missingness , for example , can result in small numbers of SNPs with complete genotypes across all samples , leading to loss of statistical power and introduction of potential bias . Imputation methods that reconstruct haplotypes using linkage disequilibrium ( LD ) patterns have become routine ways of handling missing genotypes in human genome-wide studies [16–19] . However , these methods may be compromised in low LD settings , where high rates of recombination in P . falciparum lead to decay in inter-SNP correlation within 500 bp [2 , 16] . Validating the imputation of missing genotypes in whole-genome sequences could enable population genetics methods to scale more appropriately with increasing sequences and SNPs , without having to resort to customized missing data strategies . Here we present a novel evaluation of LD-based imputation applied to a large number of global P . falciparum samples and demonstrate its utility in population genetics analysis . Two state-of-the-art methods , IMPUTE [17] and Beagle [18] , were assessed by masking known genotypes and measuring the accuracy of imputed genotypes using an allelic r2 statistic [18] . Briefly , IMPUTE is based on the LD framework of Li and Stephens [19] , with the haplotype to be imputed modelled as a sampled path through all possible underlying states defined by a hidden Markov model ( HMM ) constructed from a set of complete haplotypes . Externally inferred recombination rates determine the transition probabilities from one haplotype position to the next , while a mutation rate parameter controls the emission probabilities determining fidelity from hidden to observed states . Beagle constructs an HMM by clustering observed haplotypes at each marker position such that hidden states correspond to groups of haplotypes [18] , with each group representing haplotypes with similar transition probabilities for nearby downstream alleles . Recombination is implicit within the modelled transition probabilities and no mutation parameter is incorporated . We took advantage of the structure of IMPUTE software to test LD-based recombination rates derived from 2 high-quality parasite genetic crosses ( 7G8-Brazil × GB4-Ghana [20]; DD2-Indochina × HB3-Honduras [21] ) and a coalescent-based model of LD in LDhat software [22] . This was complemented by evaluating the empirical model of LD in Beagle software based on its haplotype clustering algorithm [18] . By completing multiple SNP datasets using these methods , we performed genome-wide scans for signatures of natural selection using EHH-type [23] and Tajima’s D [11] metrics , and a complementary analysis of parasite population structure . P . falciparum populations are geographically dispersed and undergo regional adaptation due to selection pressures from antimalarial drugs , and human and mosquito immune responses . The effect of imputation on patterns of population structure was assessed using principal components analysis ( PCA ) and pairwise population Fst [24] . All results were compared to a “complete-case analysis , ” where SNPs with missing genotypes were excluded . We demonstrate that currently available LD-based imputation methods used previously in human populations can be extended to sequence data from P . falciparum populations , with imputed genotypes reliably capturing known and potentially novel loci encoding drug resistance phenotypes and candidate vaccine antigens .
Sequencing reads from 2 Southeast Asian ( Thailand , n = 91; Cambodia , n = 253 ) and 2 African ( Gambia , n = 55; Malawi , n = 60 ) populations of P . falciparum clinical isolates [2] were aligned to the P . falciparum 3D7 reference genome ( version 3 ) [25] . A total of 85 , 967 biallelic SNPs ( “86k set” ) across these 459 isolates were identified , representing a marker density of 1 SNP every 270 bp . The 86k set represents 4 , 799 genes ( SNPs per gene: median , 4; 90th percentile , 17 ) and was dominated by markers that are monomorphic in at least 1 population ( Thailand , 50 , 087 ( 58% ) ; Cambodia , 34 , 049 ( 40% ) ; Gambia , 53 , 539 ( 62% ) ; Malawi , 50 , 287 ( 59% ) ) . The Southeast Asian populations share 26 , 551 ( 31% ) SNPs while the African populations share 24 , 089 ( 28% ) SNPs . Only 7 , 855 ( 9 . 1% ) SNPs are shared between all populations . Excluding SNPs that are monomorphic in each population , low-frequency SNPs predominate in the 86k set ( minor allele frequency ( MAF ) <5%: Thailand , 62%; Cambodia , 72%; Gambia , 45%; Malawi , 59% ) ( Fig 1 ) , and the overall MAF is low ( median 0 . 8%; 90th percentile 11% ) . Using common SNPs ( MAF≥5% ) , there was evidence that LD decayed rapidly within a few hundred base pairs in the African populations , and reached a baseline within 500 bp . Compared to the Southeast Asian populations , LD decayed more rapidly in the African populations ( Fig 2 ) , consistent with their higher recombination rates ( S1 Table ) . The estimated recombination rates are especially high in Malawi , resulting from high levels of outcrossing associated with relatively high transmission intensity and multiplicity of infection [6] . Isolates and SNPs had less than 10% of their genotypes missing , reflecting the current practice of analysing only high-quality samples ( median , 0 . 3%; 90th percentile , 3 . 1%; maximum , 8 . 9% ) and markers ( median , 0 . 4%; 90th percentile , 2 . 4%; maximum , 8 . 3% ) . The number of SNPs with complete genotypes varied by population ( Thailand , 56 , 797 ( 66% ) ; Cambodia , 47 , 482 ( 55% ) ; Gambia , 53 , 469 ( 62% ) ; Malawi , 54 , 419 ( 63% ) ; Table 1 ) . Restricting analysis to those SNPs with complete genotypes across the 459 samples led to severe data depletion , reducing the number of SNPs from 85 , 967 to 21 , 077 ( 24 . 5% ) . A strong positive correlation between SNP-wise rate of missingness and MAF ( S1 Fig ) would have excluded most common SNPs from a complete-case analysis of these 21 , 077 SNPs ( MAF: median , 0 . 4%; 90th percentile , 1 . 5% ) . This correlation could be an artefact of genotype calling from sequencing reads . In particular , because the genotypes were called using ratios of coverage , genotypes from SNPs with a higher minor allele frequency are less likely to pass the calling threshold due to more balanced coverage ratios . If the threshold was lowered , more of those genotypes are likely to be called erroneously . SNPs in chromosome 13 were chosen to compare imputation strategies , since this is the second-to-longest chromosome containing correspondingly large numbers of SNPs , and it contains the kelch13 locus recently associated with artemisinin resistance in Southeast Asia [3] . Using a leave-one-out cross-validation approach , marker genotypes were masked and the accuracy of imputed calls to the true allele was measured using an allelic correlation metric r2 ( Table 1 ) . The imputation accuracy of Beagle was consistently high ( mean r2 , 0 . 87–0 . 96 ) , reaching levels seen in a human genome-wide setting [19] . However , there was substantial variation in IMPUTE results across populations , with near-perfect mean r2 values in Gambia ( 0 . 99 ) and high values in Cambodia ( 0 . 77–0 . 84 ) , contrasting with lower values in Thailand ( 0 . 51–0 . 56 ) and Malawi ( 0 . 30–0 . 69 ) . Using a cosmopolitan panel of reference haplotypes from all 4 populations ( including 12 Vietnamese samples with complete genotypes ) , the accuracy of IMPUTE was improved in Thailand ( mean r2 , 0 . 70–0 . 71 ) and Cambodia ( 0 . 82–0 . 90 ) but degraded in Malawi ( 0 . 25–0 . 26 ) . Using the same cosmopolitan panel , the accuracy of Beagle was substantially degraded in all 4 populations including Thailand ( mean r2 , 0 . 27 ) . For IMPUTE , recombination rates inferred from LDhat were marginally more accurate than those inferred from analysing 2 genetic crosses . In Malawi , accuracy was substantially improved when using a reference panel specific to this population ( mean r2 , 0 . 30–0 . 69 ) . Using a lower mutation rate parameter ( θ = 0 . 001 ) or effective population size ( Ne = 10 , 000 ) led to only marginal improvements in accuracy in the Thai ( mean r2 , 0 . 62–0 . 64 ) , but not Gambian ( mean r2 , 0 . 99 ) population ( S2 Table ) . Rates of minor allele concordance , defined as the agreement between absolute genotype calls ( made on a threshold of ≥0 . 9 on posterior probabilities ) and true minor alleles , tracked the measured r2 closely ( Pearson’s correlation = 0 . 75 ) . The accuracy of LD-based imputation increases with MAF in the Southeast Asian populations , with most improvement occurring in the low-frequency range ( Fig 3 ) . This increasing trend is also observed in studies imputing human genotypes [16] . In Gambia , the accuracy of Beagle is generally high except for one SNP at MAF 47% with r2~0 . In contrast , the near-perfect accuracy of IMPUTE in Gambia is maintained across the allele-frequency spectrum . In Malawi , because of the limited range ( 3–23% ) of MAF values and paucity of SNPs with higher frequencies , a trend for r2 could not be clearly discerned in any of the imputation designs . Across all designs there was no systematic change in MAF post-imputation ( S2 Fig ) . Cross-validation experiments indicated that the optimal imputation accuracy of Beagle occurs when a population-specific reference panel is used , and that for IMPUTE occurs when LDhat-derived recombination rates and a cosmopolitan reference panel are used . For each imputation design , a set of 10 SNP datasets was completed by filling in missing data using sampling of posterior genotype probabilities ( Materials and Methods ) . To assess the effect of imputation strategies , we focused on selection metrics that require complete SNP datasets , namely those that detect long-range haplotypes and extended-haplotype homozygosity , both within ( |iHS| ) and between ( Rsb , a version of XP-EHH [26] ) populations . These metrics were computed on each dataset , and results pooled using a multiple imputation approach to provide overall estimates . Pearson’s r correlation of the selection metrics between Beagle- and IMPUTE-derived haplotypes was very high across the 4 populations , for both |iHS| ( range , 0 . 93–0 . 98 ) and Rsb ( range , 0 . 98–0 . 99 , using Malawi as the reference population ) . IMPUTE-derived haplotypes systematically resulted in higher Rsb values compared to Beagle-derived haplotypes ( S3 and S4 Figs ) . The correlation between selection metrics from Beagle-imputed versus complete-case genotypes was much lower ( e . g . , |iHS| r<0 . 60; S5 and S6 Figs ) . Given the arguably superior performance of Beagle and its ease of use , we proceeded by comparing the top selection hits ( overall 1% of threshold across populations , |iHS|>2 . 45 , Rsb>4 ) between Beagle-imputed and complete-case haplotype analyses . Evidence for selective sweeps due to drug pressure or other mechanisms was investigated using the |iHS| and Rsb metrics applied to Beagle-imputed haplotypes . Intra-population analysis revealed signals downstream of crt , including cg1 and cg2 , in the African populations , but not in the Southeast Asian populations in which mutant alleles are already fixed ( Fig 4 and S3 Table ) . Other strong |iHS| hits included genes encoding vaccine candidates ( e . g . , ama1 and trap ) and other membrane and surface proteins ( e . g . , clag2 and msp4 ) . In contrast , analysis of complete-case haplotypes only identified ama1 ( S4 Table ) . Inter-population analysis using the Rsb metric has the potential to detect positively-selected alleles that have already achieved fixation . With Malawi as the reference population , crt and neighbouring cg1 and cg2 were strongly identified across the other 3 populations ( Fig 5 and S5 Table ) . Whilst sulfadoxine-pyrimethamine resistance genes were not directly identified , there were strong signals in regions proximal to dhfr ( e . g . , PF3D7_0417400 , S7 Fig ) and dhps ( e . g . , PF3D7_0809800 , S8 Fig ) in Southeast Asian populations . This phenomenon of signal ‘shifting’ is due to the beneficial allele sweeping up in parallel in the Malawian reference population , giving rise to long-range haplotypes that attenuate the inter-population differences in the EHH at regions close to the SNP conferring drug resistance . Rsb analysis also confirmed positive selection in ama1 and trap in Southeast Asian populations . Analysis of complete-case haplotypes failed to detect crt , dhfr , and dhps effects ( S6 Table ) . Rsb analysis of these haplotypes did , however , confirm positive selection in ama1 and trap in Southeast Asian compared to Malawian populations . Importantly , this analysis also identified positive selection in kelch13 [3] in Cambodian compared to Thai populations ( S7 Table ) , as well as in 2 other loci: a ~600kb region downstream of kelch13 ( Fig 6 ) ; and PF3D7_0104100 , in a region upstream of pfubp1 [5] , detected previously in a selecton scan of Kenyan parasites [5] . In this instance , analysis of complete-case haplotypes also managed to identify a signal in kelch13 ( S7 Table ) , indicating that the sweep is young and the signal has not yet been lost through recombination . The complete-case haplotype analysis also detected pfs45/48 and pfs47 that mediate evasion of the mosquito immune system , and are a signature of highly-structured populations resulting from inter-continental differences in the prevalence of Anopheles vector species [27] . Using Beagle-imputed haplotypes , Tajima’s D was calculated by population for each of the 3 , 517 ( of 4 , 799 ) genes with 3 or more SNPs . The vast majority of Tajima’s D values were negative ( Thailand , 85%; Cambodia , 89%; Gambia , 86%; Malawi , 93% ) , indicating a demographic history of recent population expansion [6 , 28 , 29] or purifying selection . The malaria parasite life cycle itself can also lead to a skewed frequency spectrum toward low frequency alleles and negative Tajima’s D [29] . Amongst the genes with high Tajima’s D values ( >1 ) based on at least 10 SNPs ( Table 2 ) , several encode well-described targets of acquired immunity such as ama1 ( all populations , 1 . 16–1 . 79 ) and msp1 ( Southeast Asia , 2 . 23–2 . 29 ) . Other genes previously found to have high Tajima’s D values [7] were also identified , including eba175 ( all populations , 1 . 02–1 . 55 ) , dblmsp ( non-Malawi , 1 . 22–2 . 11 ) , and clag2 ( non-Malawi , 1 . 35–1 . 71 ) . Both ama1 and msp1 have been previously identified to be under positive or balancing selection in certain parasite populations [5 , 29] . Using the set of unimputed SNPs , nearly all of the 3 , 517 Tajima’s D values were negative ( Thailand , 99 . 5%; Cambodia , 100%; Gambia , 99%; Malawi , 100% ) , reflecting the exclusion of common SNPs and inability to calculate pairwise nucleotide differences between haplotypes in a population with missing genotypes . The exclusion of common SNPs may also lead to an overestimate of population expansion . Principal components analysis using both unimputed or imputed data revealed similar well-defined population structure , namely clustering firstly at a continental level ( S9 Fig ) , and secondarily at regional and population levels . The extent to which SNPs drive these inter-population differences was quantified using the Fst differentiation metric . As expected , greater mean Fst values were attained between ( range: unimputed , 0 . 260–0 . 273; imputed , 0 . 259–0 . 267 ) than within ( range: unimputed , 0 . 035–0 . 057; imputed , 0 . 040–0 . 058 ) continents ( S8 Table ) . As the Fst implementation was SNP- rather than haplotype-based , there was little appreciable difference between unimputed and imputed results ( P>0 . 05 ) , further supporting the robustness of population structure analysis of datasets with low frequencies of missing genotypes .
The ability to monitor for signatures of selection in P . falciparum is being revolutionised by whole-genome sequencing of large numbers of global clinical isolates . One unintended consequence of this effort , however , is that sequencing greater numbers of parasite genomes will increase the number of missing genotypes , presenting new challenges to haplotype-based population genetics analysis that requires complete SNP datasets . Imputation approaches have been successfully applied to human genomes , but high rates of recombination and rapid LD decay challenge their application to P . falciparum genomes . To help meet these challenges , our study shows that readily available software ( Beagle and IMPUTE ) produce accuracy levels comparable to those in human studies , and identify known P . falciparum loci involved in drug resistance and immune evasion . Application of Beagle with a population-specific panel of reference haplotypes provided consistently high levels of imputation accuracy across 4 parasite populations , suggesting that its estimation of local haplotype clusters reliably reproduces LD structure in each population . In contrast , accuracy with IMPUTE varied between populations using a similar population-specific referencing strategy . Recombination maps were derived from genetic crosses of parental strains from Honduras , Indochina , Brazil , and Ghana . These are unlikely to represent haplotype configurations that accurately characterise LD structure in other populations , and therefore led to lower accuracy of samples from geographically-distant countries ( e . g . , Thailand ) in our study . Inferring recombination maps from population haplotypes via LDhat marginally improved the performance of IMPUTE . There were differences in recombination rates between inferred and experimentally derived maps ( S10 Fig ) . The near-perfect accuracy in Gambian samples observed for IMPUTE was replicated consistently across different recombination maps , due in part to lower levels of missing genotypes , but this was the exception . Adopting an ancestrally-diverse reference or cosmopolitan panel within IMPUTE , however , significantly increased accuracy across the allele-frequency spectrum . Using a cosmopolitan panel unconstrains the IMPUTE algorithm to choose the most similar set of haplotypes , including those where the parasites of origin may be genealogically distant [30] . However , use of such a panel worsened Beagle’s performance in all populations , due in part to its less flexible haplotype clustering and LD modelling on all haplotypes supplied [18] . While useful to test the validity of LD models via recombination maps derived from different sources , the design of the IMPUTE software is such that LD is modelled independently of missing genotypes using a complete reference panel [17] , suggesting that the sporadic missingness of sequencing data may be problematic to the construction of such a model . In addition , the necessity of many free parameters required as inputs made it a less robust imputation strategy for P . falciparum . Newer methods of LD inference such as LDhelmet [31] may improve the underlying model of recombination; however , IMPUTE is intrinsically not very amenable to imputing sporadically missing genotypes . In contrast , Beagle infers LD by jointly modelling the missing and observed genotypes . Rather than making hard genotype calls , genotype probabilities inferred from the Beagle and MACH [32] methods could improve the handling of uncertainty at the SNP discovery phase and potentially enable more SNPs to be imputed and used in downstream analyses . The top gene hits using EHH-based metrics computed from imputed haplotypes suggest that imputation is successful in reproducing signatures of positive selection , even when large numbers of haplotypes with initially extensive missingness are used . As in previous haplotype scans for signatures of positive selection in P . falciparum [5 , 33–36] , we confirmed the crt signal using the inter-population metric Rsb . Signals from the intra-population metric |iHS| ( with MAF≥5% ) were likely to miss evidence of near or full selective sweeps , as in Southeast Asia where beneficial crt alleles and their associated long-range haplotypes have reached fixation . More-recent selective sweeps were observed in regions close to the sulfadoxine-pyrimethamine ( SP ) resistance genes dhfr and dhps in the Thai , Cambodian , and Gambian populations . Unlike chloroquine resistance that spread out of Southeast Asia , SP resistance arose independently across continents . The Rsb metric highlighted the known independent sweep in dhps in Malawi , leading to signal ‘shifting’ in other populations . This metric also identified the artemisinin resistance gene kelch13 within a ~600kb region in Cambodian parasites , with Thai parasites as the reference population . This study is the first to detect kelch13 using a phenotype-independent , whole-genome scan for signatures of positive selection . Since kelch13 mutations have evolved only recently , this finding suggests that global surveillance of whole-genome P . falciparum sequence data may successfully detect entirely new forms of resistance to future drugs and vaccines . Analysis of complete-case haplotypes also detected the kelch13 signature of selection , likely because this gene had relatively high coverage ( 1 . 5-fold greater than the genome average ) . However , these haplotypes could not validate known drug resistance loci and , overall , their correlation with Beagle-imputed results was low . The absence of known loci was primarily due to exclusion of disproportionately more-common SNPs with missing genotypes . EHH-based methods depend on accurately measuring LD decay around a core SNP . There was a trade-off between inflation of EHH due to lower allele frequencies , and deflation of EHH due to a 75% reduced marker density ( i . e . , the closest markers are less likely to be in LD ) . A shift in allelic-frequency spectra towards rare alleles inflates negative Tajima’s D values , indicating directional selection . By applying a Tajima’s D approach to the imputed data , genes under balancing selection as putative antigenic determinants were identified . For example , ama1 , msp1 , and msp7 encoding well-characterized parasite antigens were detected in multiple populations , although msp3 . 8 and trap previously detected in Malawi [6] and Gambia [37] , respectively , did not exceed the stringent Tajima’s D threshold of 1 . Complete-case genotypes failed to produce robust hits , as ≥99% of Tajima’s D values were negative across all populations . The skew towards low-frequency SNPs in the complete-case dataset biases the nucleotide diversity and therefore the Tajima’s D statistic . Similarly , the effective sample size is smaller than the analysable sample size after imputation , and may bias variation estimation and Tajima’s D . Some loci thought to be under balancing selection were found to be under recent positive selection , highlighting that some loci may be under dual selective pressures depending on where and when parasite samples are collected . Although such loci have signatures similar to incomplete sweeps of positive selection [38] , evidence for dual selection pressures on ama1 and trap corroborated similar findings in other populations [5 , 18] . In summary , we have outlined a viable strategy for imputing missing genotypes in large numbers of global P . falciparum isolates , taking into account differences in LD and population structure . Specifically , we recommend using Beagle as the preferred algorithm for imputing whole-genome sequences , and population-specific haplotypes as references . This approach may be useful for imputing genome data from other Plasmodium spp . , with resultant datasets being more complete , better able to identify evidence of natural selection by antimalarial drugs and host immunity , and more likely to lead to new insights and applications for disease control .
Sequencing reads from 3 Southeast Asian ( Cambodia , n = 253; Thailand , n = 91; Vietnam , n = 12 ) and 2 African populations ( Gambia , n = 55; Malawi , n = 60 ) of P . falciparum clinical isolates [2] were aligned to the P . falciparum 3D7 genome ( version 3 ) using smalt ( described in [2] ) . For all samples the average coverage across the whole genome was at least 35-fold . Variants were called using samtools [39] and vcftools [40] with default settings . Genotypes at SNP positions were called using ratios of coverage , with heterozygous calls converted to the majority genotype on a 70:30 or greater coverage ratio [4] . SNPs that reflected possible multiplicity of infection were removed as the imputation methods under evaluation cannot tolerate haploid and mixed genotypes simultaneously . Low-coverage var and subtelomeric regions ( S9 Table ) were discarded . The 85 , 967 biallelic and non-singleton SNPs with less than 10% missing genotypes were retained , which is approximately the inflection point beyond which SNPs experience a rapid increase in missingness ( S11 Fig ) . Allowing more missing data by relaxing the threshold would lead to a decrease in imputation accuracy [17] . Performance of the imputation methods was evaluated with a leave-one-out cross-validation approach using markers on chromosome 13 , where 215 ( ~2% ) SNPs were masked in each haplotype and imputed using the others as references . This procedure produces an accuracy measure based on the squared Pearson correlation ( allelic r2 ) between the true and imputed allele probability for each marker [18] . To prevent sampling bias from choosing only markers with completely known genotypes , each population panel was split into several subpopulations of 25–30 isolates each , and independent rounds of cross-validation were performed on a distinct set of markers in each subpopulation . This procedure introduced representation of markers at varying levels of missingness , and increased the total number of validated markers . A concordance rate between true and best-guess genotypes on a calling threshold ≥0 . 9 was also calculated . Only the concordance of minor alleles is considered to mitigate the possibility of near-monomorphic SNPs inflating overall concordance rates . Imputation software established primarily for human genotype analysis were adapted to enable their application to parasite sequence data . Beagle v3 . 3 [18] does not require recombination rates to inform its LD model and processes sporadically missing genotypes , but only imputes diploid genotypes . Therefore , we duplicated all positions to make sequences homozygous diploid before each imputation round , then heterozygous imputed genotype probabilities were allocated equally between the 2 alternative alleles when computing allelic dosage . For IMPUTE v2 . 0 , controls allowing haploid imputation ( -chrX , -use_prephased_g , -known_haps_g ) were activated . Previously published recombination maps from 2 high-quality genetic crosses between laboratory clones GB4x7G8 [6 , 21 , 38 , 41] and HB3xDD2 [6 , 22 , 38 , 41] were used to supply 2 sets of experimentally informed rates . These two maps were considered separately , and not combined . An alternative set of population recombination rates ( 2Ner ) was inferred using the coalescent-based LDhat program , interval [42] . Pre-supplied likelihood lookup files with a population mutation rate ( θ ) value of 0 . 01 were used . The effective population size ( Ne = 30 , 000 ) was derived by re-scaling the compound map distance for Gambian isolates inferred from interval , to match the empirical map distance estimated from the GB4x7G8 cross . The scaling factor was treated as a pseudo-Ne used to adjust compound rates uniformly on the other populations . To test the robustness of assuming the above parameter values for θ and Ne , we computed accuracy using different values for one African ( Gambia ) and one Southeast Asian ( Thailand ) population ( S2 Table ) . Multiple haplotype sets were sampled from the marginal posterior genotype distribution under the best-performing imputation designs . This procedure allowed the computation of population genetics metrics under a multiple imputation-like schema , accounting for uncertainty in the true identity of genotypes . Ten SNP datasets were completed for each imputation strategy through sampling from posterior genotype probabilities , and relevant population genetics metrics at each SNP or gene were computed and averaged across the sets . Two metrics were used to infer recent positive selection , the integrated haplotype score ( iHS ) and Rsb . iHS is the standardized log ratio of the integrated extended-haplotype homozygosity ( EHH ) between the ancestral and derived allele at a core SNP [43] , capturing evidence of unusually long haplotypes surrounding a particular allele within the population . The ancestral and derived alleles were defined to be the major and minor allele at Malawian SNPs , as this population is ancestrally ancient [6 , 16] and the rapid LD decay made it a natural reference point for detecting extended haplotype homozygosity in other populations . Absolute values of iHS were used to capture long haplotypes centred around either type of allele , using a minimum MAF filter of 5% . Rsb is the standardized log ratio of the integrated site-specific EHH between populations at a core SNP , where site-specific EHH refers to the weighted average of EHH surrounding a core SNP according to squared-allele frequencies [26] . Because site-specific EHH does not require markers to be polymorphic within the population , it can detect selective sweeps for alleles that have risen to fixation . Both iHS and Rsb were calculated using the scan_hh program in the R package rehh [44] , which excludes markers with any missing data . The allele frequency-based Tajima's D approach [11] was used to detect balancing selection at each gene , with greater evidence from increasing values from zero . Whilst different populations have different demographic histories , it is difficult to account for this and establish multiple thresholds for the different null distributions of Tajima’s D . To negate any confounding effects of population expansion and difficulties in establishing significance levels , we report genes with Tajima’s D values in excess of one [6] . The function neutrality . stats from the R package PopGenome was used [45] . Parasite population structure was analysed using principal components analysis , implemented via the classical multidimensional scaling method using the function cmdscale in R . In addition , pairwise population Fst with 95% bootstrap confidence intervals was computed using the function stamppFst in the R package StAMPP [46] . | Characterizing genetic diversity and function in Plasmodium falciparum , including identifying determinants of emerging drug resistance , is crucial to informing public health strategies to contain and eliminate this malaria parasite . The lack of a robust framework to handle missing P . falciparum genotypes arising from next-generation sequencing efforts , impedes genome-wide methods that depend on complete genotype information , and often leads to analysis that discards entire regions of the genome . This study is the first to evaluate the performance of missing data imputation or “filling in” in the P . falciparum genome , where the correlation between genetic markers is generally lower than in the human genome . We considered 86k markers in 459 clinical isolates from 4 malaria-endemic populations of Africa and Southeast Asia . Although low genotype missingness per SNP ( <10% ) results in complete datasets for only 25% of SNPs , imputation is accurate . This finding is corroborated by the ability of imputed haplotype analysis to recover several well-established vaccine candidates and drug resistance loci , including kelch13—a recently-validated gene involved in artemisinin resistance . Our work demonstrates that imputation can assist the application of genome-wide methods to identify the determinants of P . falciparum diversity , including those involved in drug resistance , immune evasion , and host virulence . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2015 | Imputation-Based Population Genetics Analysis of Plasmodium falciparum Malaria Parasites |
Climate change disrupts ecological systems in many ways . Many documented responses depend on species' life histories , contributing to the view that climate change effects are important but difficult to characterize generally . However , systematic variation in metabolic effects of temperature across trophic levels suggests that warming may lead to predictable shifts in food web structure and productivity . We experimentally tested the effects of warming on food web structure and productivity under two resource supply scenarios . Consistent with predictions based on universal metabolic responses to temperature , we found that warming strengthened consumer control of primary production when resources were augmented . Warming shifted food web structure and reduced total biomass despite increases in primary productivity in a marine food web . In contrast , at lower resource levels , food web production was constrained at all temperatures . These results demonstrate that small temperature changes could dramatically shift food web dynamics and provide a general , species-independent mechanism for ecological response to environmental temperature change .
The ocean is a dynamic part of the global climate system . The temperature of the sea surface , where almost 50% of the world's primary productivity occurs [1] , varies regionally as the result of changing surface air temperatures , currents , and upwelling of deeper water . Though links between climate conditions and pelagic food web productivity and structure have long been of interest to scientists [2] , effects of physical conditions on secondary and tertiary productivity ( hereafter: consumer productivity ) have seemed too context dependent to allow general predictions [3]–[5] . The prevailing conceptual framework for understanding effects of ocean temperature on food webs is based on the view that consumer production is predominantly controlled indirectly by temperature effects on primary production [6] , [7] . According to this model , increased primary productivity and net autotrophy also increase CO2 uptake of the whole food web [8] , [9] . Yet recently developed metabolic theory and a meta-analysis indicate that heterotrophic ( respiration-limited ) metabolism is more sensitive to changing temperature than autotrophic ( photosynthesis-limited ) metabolism and production ( Figure 1A ) [9] , [10] , suggesting stronger consumer-driven control with warming . Greater consumer control of primary production would lead to increased heterotrophy and less phytoplankton standing stock ( Figure 1B-ii ) . In either model , the response of food web productivity and structure to changing environmental temperature may be determined by general processes and not the specific responses of component species , and thus could represent a critical step forward in efforts to forecast the impacts of climate change on ecological communities [11] , [12] . Temperature-driven shifts in food web productivity and structure are limited ultimately by resource availability , and therefore must be considered in realistic nutrient supply contexts ( Figure 1B ) [5] . In the ocean , the same physical processes that drive temperature patterns also influence resource availability . Temperature-driven stratification isolates surface waters from cool , nutrient-rich deeper water , and because biological productivity at the sunlit surface depletes available nutrients , temperature and nutrient supplies are usually negatively correlated [5] . Nutrient limitation directly constrains primary production , while metabolic responses to temperature influence both photosynthetic and respiratory processes , and thus primary and consumer production . The metabolic effects of temperature therefore should be different and complementary to constraints imposed by resource availability . To understand the combined effects of temperature and resource availability on food web biomass ( gC L−1 ) and productivity ( gC L−1 yr−1 ) , whole food web responses to variation in both factors need to be assessed . Using a coastal pelagic food web of phytoplankton producers and bacterial and zooplankton ( >63 µm ) consumers , we experimentally tested the effects of non-lethal temperatures and resource availability on food web structure ( biomass allocation among trophic levels ) and biomass standing stock ( gC L−1 ) . We assembled food webs in outdoor microcosms in a factorial experiment with four temperature levels ( ambient , +2 , +4 , and +6°C ) and two resource levels ( nutrient additions and controls ) ( Table 1 ) . Treatment levels mimic local estuary conditions during springtime warming and dry versus storm events causing riverine inputs of elevated nutrient concentrations ( Figure S1 ) . We measured effects of temperature and nutrient treatments on standing stocks of primary and secondary producers and on rates of primary productivity . Initial microcosm conditions included known amounts of zooplankton , phytoplankton , and bacteria ( Figure 2 ) collected from the Bogue Sound Estuary at the University of North Carolina's Institute of Marine Sciences ( IMS ) in Morehead City , North Carolina .
We found that small increases in temperature ( Table 1 ) shifted food web structure toward greater heterotroph biomass relative to autotroph biomass ( H/A ) ( Figure 2A ) . This shift is consistent with predictions based on differential temperature scaling of respiration- and photosynthesis-limited metabolism ( Figure 1 ) [9] , [13] , [14] . Differential temperature scaling implies that organismal processes such as resource use , growth , and reproduction rates scale differently with temperature for heterotrophs and autotrophs [9] , [14] . Consequently , increased grazing pressure with temperature dramatically reduced standing phytoplankton biomass in spite of increased per capita primary productivity ( as approximated by the maximum photosynthesis per unit chlorophyll biomass , PMB , Figure 3 , Table S1 ) . Stronger consumer effects and greater consumer biomass were driven by higher density , and not increased individual size or a shift in the relative abundance of species ( Figure S2 ) . This pattern is consistent with the hypothesis that temperature affected change on a metabolic , individual level rather than via competitive exclusion or other species interactions . Shifts in food web structure with warming were accompanied by a decrease in overall biomass ( Figure 2E , Table 2 ) . The decline in total biomass is consistent with stronger consumer control of food web structure with warming [14] , [15] , reflecting a direct effect of temperature on consumers and a disproportionate increase in grazing relative to primary production . Conversion of phytoplankton into consumer biomass is inefficient ( ∼10% [16] ) , so as consumers represent a greater proportion of the food web biomass , total biomass must decline . If instead the predominant influence of temperature on consumer productivity had been mediated indirectly by increased primary productivity , total food web biomass would have increased ( Figure 2B-i ) . Although food web structures with reduced relative primary producer biomass are thought to be unstable , Carpenter et al . ( 2001 ) [15] showed that such a top-heavy food web structure can be sustained over time in natural pelagic lake food webs ( i . e . , at least 5 years ) . There was a strong interaction between resource availability and temperature effects on food web structure and total biomass . Nutrient addition allowed food web structure ( H/A ) to increase with warming and led to greater total food web biomass that declined with warming ( Figure 2 ) . In contrast , in the nutrient control treatments , resource availability constrained primary productivity at all temperatures ( Figure 3 ) , limited total standing biomass , and reversed the temperature-induced increase in H/A at the highest temperature . Low H/A at the highest temperature probably reflects exhaustion of resources at the highest productivity rates . In sum , resource availability determined whether food web structure was more strongly influenced by resources or by consumers ( Figure 1B ) . If this experimental system is representative of effects of environmental warming , the interaction between nutrient supply and temperature suggests that in nutrient-poor regions , food webs may be more resilient to warming because consumer production is limited by resource availability , while in nutrient-rich regions small amounts of warming may have dramatic effects on trophic structure , primary productivity , and standing biomass . Food web experiments in microcosms are a necessary but imperfect approximation of natural conditions . Microcosm experiments allow manipulation of environmental factors that would be impossible in the field while allowing natural feeding interactions , behaviors , and population growth processes to occur . These advantages undoubtedly enhance our understanding of fine-scale biological dynamics in pelagic ecosystems . Nonetheless , small microcosms impose several limitations on the broad interpretation of their results . For example , evaporation at warmer temperatures increased salinity in our microcosms ( Table 1 ) . Reduced concentrations of dissolved oxygen and carbon dioxide are also associated with warmer temperatures and likely varied naturally in our microcosms . Though these factors can influence productivity , their effects are negative and small over the experimental temperature range relative to the strong positive effects of temperature [17] . In addition , it is possible that the importance of consumer control was amplified in our experimental microcosms . For example , small experimental systems with relatively homogenous environments can facilitate foraging and reduce refuges for resources . Nonetheless , top-down determination of food web structure and dynamics has been documented in large-scale aquatic ecosystems [15] , [18] and may become more important in a warming environment .
Temperature is known to influence food web structure [13] , [19] , [20] , and such findings have generally been attributed to differential effects of resource limitation across trophic levels , or the specific effects of temperature on consumers or producers [13] , [19] . Our experiments demonstrate that temperature alone can shift food web structure and change total standing biomass . Furthermore , biogeographic trends towards net heterotrophy in warmer climates in open ocean pelagic food webs [9] , [21] and patterns observed in spring bloom dynamics , rocky intertidal systems , grasslands , and forests [20] , [22]–[25] are consistent with differential metabolic scaling across trophic levels , though this mechanism has been invoked and tested in just one of these cases [9] . By explicitly testing the hypothesis based on metabolic theory in the context of food web ecology , we have for the first time experimentally validated the prediction that universal temperature constraints on individual metabolism can lead to general responses at the community level [11] , [26] . The interaction between effects of temperature and nutrient availability observed in these experiments deepens our understanding of food web responses to changing climate conditions . In pelagic marine ecosystems , projected increases in stratification imply that negative correlations between nutrient availability and temperature will intensify in many regions [5] , [27] . This pattern occurs on very broad geographic scales ( i . e . , cold temperate or polar systems relative to tropical systems ) , on smaller scales within oceans and seas ( i . e . , the North Sea [19] ) , and over time within a single region [4] . According to theory supported by our experimental results , small increases in sea surface temperature should cause small declines or no change at all in primary productivity and standing stocks in nutrient-poor systems such as stratified areas with a shallow thermo- or pycnocline . Under such conditions , nutrient limitation would constrain consumer productivity and biomass stocks ( Figure 1B-iii ) , and could even lead to reduced consumer biomass with warming due to increased respiratory costs that exceed available primary production . In contrast , when nutrients are plentiful , as in upwelling or well-mixed systems , warming should increase productivity leading to increased biomass production at higher trophic levels , shifted food web structure , and stronger consumer control of phytoplankton standing stock ( Figure 1B-ii ) . The importance of temperature scaling of food web structure for fisheries productivity and food webs in aquatic ecosystems depends on the contemporary food web structure . In nature , most food webs include consumers at trophic levels higher than the zooplankton used in our microcosm experiments . In more complex food webs , temperature-driven intensification of consumer control could strengthen a trophic cascade , causing increased phytoplankton biomass as a result of indirect effects of increased consumption by carnivores . Alternatively , if consumer biomass has been severely reduced due to overfishing , direct effects of differential temperature scaling across trophic levels may be difficult to detect and indirect effects of increased primary productivity may be most apparent ( Figure 1B-i ) [3] . Ocean warming or cooling influences marine ecosystems in a variety of ways . For example , together with associated changes in physical properties such as vertical stratification and ice cover , warming has shifted species composition and altered the timing of seasonal spawning and spring bloom events [5] , [28] , [29] . The ramifications of these changes can be severe for some species and mild for others , causing mismatch between interacting species [6] , [20] , [28] . Temperature scaling of food web properties , however , is a general response to temperature change that should occur regardless of species composition [9] , [10] . This mechanistic response can be incorporated into predictions of ecological variation , thus providing one of the few general models for ecosystem change with geography or climate . The conceptual framework outlined here reinforces predictions that effects of climate change on ecosystem processes will vary among regions [7] , [30] . Future warming will likely increase secondary productivity and fish harvests in nutrient-rich regions , but may cause little change in more stratified , oligotrophic systems . These are not paradoxical responses , and the general effects of temperature in different nutrient contexts explain why different responses to warming can occur within the same ecosystem . Implications of temperature effects on food webs for the ocean's role in carbon cycling are unclear , due in part to the mosaic of nutrient-rich and nutrient-poor regions of the world's oceans , and to temperature-driven shifts in the threshold dividing net heterotrophy from net autotrophic ( carbon sinks from carbon sources ) [9] . Nonetheless , small degrees of warming may have predictable broad scale consequences for the productivity and structure of aquatic ecosystems .
Food webs were maintained in 4-L translucent plastic microcosms ( n = 5 ) in outdoor water tables at IMS from April 23 to May 1 , 2008 . Pilot experiments indicated that 8 days were sufficient to allow zooplankton population growth without exhausting water quality . We maintained temperature treatments in a blocked design with temperature blocked by water table ( Table 1 ) . Temperature treatments were significantly different ( one-way ANOVA with temperature as a continuous variable: df = 1 , F = 567 . 72 , p<0 . 001 ) , and water table did not alter the treatment effects ( comparison of nested linear models using likelihood ratio tests indicated no improvement by including a water table term: p>0 . 952 ) . Temperatures were monitored regularly using a hand thermometer and continuously using ibutton Thermochron dataloggers ( Dallas semiconductor , Dallas , Texas , USA ) . Nutrient addition and control replicates were randomly arranged in water tables . Plexiglass and one layer of window screen were placed several inches above microcosms to block UV radiation , minimize evaporation , and reduce light levels to those similar to 0 . 5–1 . 0 m depth ( approximately 900 µM photons/m2/s midday on a sunny day ) , while still allowing unhindered gaseous exchange with the atmosphere . Each microcosm received air through an air stone to maintain oxygen levels and water mixing . Phytoplankton biomass was estimated by quantifying chlorophyll a concentrations in 50 mL aliquots of each replicate . Nutrient ( NH4 , PO4 , NOX and total nitrogen ( TN ) ) concentrations were quantified using the filtrate from the same water samples used to estimate phytoplankton biomass . Zooplankton were sorted from water remaining in the microcosm after other sampling ( 2 , 768 mL ) using a 63 µm mesh and preserved in 4 . 5% sucrose Formalin . In the laboratory , zooplankton were counted and identified to lowest taxonomic level possible at 40× magnification . Carbon biomass was estimated by converting from chl a , ash free dry weight , and visual counts for phytoplankton , zooplankton , and microbes , respectively ( Text S1 ) . Final maximum primary productivity was estimated using photosynthesis versus irradiance ( P-I ) relationships for ambient , +2 , and +6°C . Maximum photosynthesis per unit chlorophyll biomass ( PMB ) and the initial slope of the P-I curve ( ±95% confidence intervals ) were calculated based on estimation of radioactive carbon uptake at each treatment level . Phytoplankton samples were collected from microcosms and spiked with 14C-bicarbonate ( Amersham ) to a final concentration of 0 . 8 µCi mL−1 and incubated for 45 minutes at varied irradiances ( Text S1 ) . Effects of temperature and nutrient levels on response variables were analyzed using a two-way ANOVA . Biomass data were log-transformed prior to analysis to meet the assumptions of ANOVA . All statistical analyses were performed in R ( v . 2 . 7 . 0 ) . P-I curve fitting was performed in SAS . | Humans rely on marine ecosystems for economic and nutritional sustenance—including about 16% of animal protein consumed by humans—making it especially important for natural scientists , economists , conservationists and long-term policy planners to understand how climate change is likely to affect oceanic food webs . Yet the general effects of warming on food web productivity are completely unknown . The productivity of consumers ( such as zooplankton ) , in food webs is determined in large part by their metabolic rates and the availability and productivity of their limiting metabolic resources . A general theory relating food web dynamics to temperature suggests that fundamental differences between consumers and primary producers ( such as phytoplankton ) may lead to predictable shifts in their relative abundance and productivity with warming . We experimentally tested the effects of warming on food web structure and productivity under two resource supply scenarios . Our results show that warming alone can strengthen the role of consumers in the food web , increasing consumer biomass relative to producer biomass , and reducing the total biomass of the food web despite increases in primary productivity . In contrast , when resources were less available , food web production was constrained at all temperatures . These results demonstrate that small changes in water temperature could drive dramatic shifts in marine food web structure and productivity , and potentially provide a general , species-independent mechanism of ecological response to climate change . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"ecology/global",
"change",
"ecology",
"ecology/community",
"ecology",
"and",
"biodiversity",
"ecology",
"ecology/marine",
"and",
"freshwater",
"ecology"
] | 2009 | Warming and Resource Availability Shift Food Web Structure and Metabolism |
Buruli ulcer ( BU ) , caused by Mycobacterium ulcerans is a chronic necrotizing skin disease . It usually starts with a subcutaneous nodule or plaque containing large clusters of extracellular acid-fast bacilli . Surrounding tissue is destroyed by the cytotoxic macrolide toxin mycolactone produced by microcolonies of M . ulcerans . Skin covering the destroyed subcutaneous fat and soft tissue may eventually break down leading to the formation of large ulcers that progress , if untreated , over months and years . Here we have analyzed the bacterial flora of BU lesions of three different groups of patients before , during and after daily treatment with streptomycin and rifampicin for eight weeks ( SR8 ) and determined drug resistance of the bacteria isolated from the lesions . Before SR8 treatment , more than 60% of the examined BU lesions were infected with other bacteria , with Staphylococcus aureus and Pseudomonas aeruginosa being the most prominent ones . During treatment , 65% of all lesions were still infected , mainly with P . aeruginosa . After completion of SR8 treatment , still more than 75% of lesions clinically suspected to be infected were microbiologically confirmed as infected , mainly with P . aeruginosa or Proteus miriabilis . Drug susceptibility tests revealed especially for S . aureus a high frequency of resistance to the first line drugs used in Ghana . Our results show that secondary infection of BU lesions is common . This could lead to delayed healing and should therefore be further investigated .
Buruli ulcer ( BU ) caused by Mycobacterium ulcerans is a necrotizing skin disease that affects mainly impoverished communities in Western and Central Africa . It is the third most common mycobacterial disease of humans after tuberculosis and leprosy . BU lesions are characterized by extensive necrosis and minimal pain and inflammation [1] , [2] . The pathogenesis of the disease is believed to be initiated by the inoculation of M . ulcerans into the subcutaneous layer of the skin , which may be facilitated by trauma or an insect vector . Most BU lesions are found at the extremities and contain extracellular clusters of acid-fast bacilli ( AFB ) in the subcutaneous fat tissue . The incubation period seems to be highly variable , and has been estimated to range from two weeks to three years , with an average of two to three months [3] . The disease begins typically as a painless nodule under the skin and gradually enlarges and erodes through the skin surface , leaving a well-demarcated ulcer with a necrotic slough in the base and widely undermined edges [3] , [4] . Traditionally , the mainstay treatment of BU was surgical removal of infected tissues followed by skin grafting [1] . This led to long hospital stays with the accompanied social problems of losses of school time by children and a large economical burden directly and indirectly to the affected families . Since 2006 , after a pilot study in Ghana , the first line treatment of BU is SR8 ( eight weeks of streptomycin daily injections and oral therapy with rifampicin ) [5]–[7] . This has reduced surgery to an adjunct procedure in BU management . The general perception is that this treatment modality will reduce the length of stay in health facilities , since it removes the fear of surgery and encourages early reporting to the formal health sector for treatment . SR8 makes a decentralization of treatment possible , since staff of peripheral health facilities can administer streptomycin injections . The pathogenesis of BU is mediated mainly by a polyketide derived macrolide toxin , named mycolactone , with potent tissue necrotizing [8] and immunosuppressive activities [9] , [10] . Mycolactone produced by clusters of M . ulcerans leads to the destruction of the surrounding soft skin tissue and to the formation of devitalized , avascular tissue and ‘necrotic slough’ at the wound bed , which is very characteristic of BU [11] . The necrotic tissue could provide an ideal medium for bacterial growth and may disturb and delay wound healing . While there is a popular belief that secondary infections of BU lesions are rare , because mycolactone has antimicrobial activities , there is no published evidence base for this . It is controversial , whether bacteria present in wounds contribute to delays in wound healing , because wounds generally harbor transient microorganisms ( contamination ) [12] . The surfaces of wounds have microbial populations at each stage of healing and some of the bacteria may be involved in mutually beneficial relationships with the host preventing more virulent organism from infecting deeper tissues . Such beneficial organisms include coagulase negative Staphylococcus and Corynebaceria species [12]–[14] . These contaminating organisms are derived from the normal flora of the surrounding skin , mucous membranes or from external environmental sources . Usually the immune defense mechanisms of the host can contain these contaminants with no harm and negative consequence to wound healing . However , some of the contaminating organisms can also go on to colonize , massively multiply and delay wound healing . Only when a critical concentration of these microorganisms is reached , signs of infection including erythema , pain , increase in temperature , odor and discoloration of granulation tissue are observed . Therefore assessment of wound infection has to be based both on the density of microorganisms as well as on the presence of specific pathogenic species [15] , [16] . Staphylococcus aureus , Pseudomonas aeruginosa , and beta-hemolytic streptococci are regarded as primary indicators for a delayed healing and infection in both acute and chronic wounds . Bacterial loads exceeding 106 colony forming units ( CFU ) /g of tissue or tissue fluid , accumulations of pus cells and presence of specific pathogenic organism are being used as indicators for wound infection in contrast to wound contamination [16]–[19] . Factors predisposing a wound to infection include the non-observance of principles of good hygienic procedures during dressing and the presence of necrotic tissue or slough within the wound margin [13] , which is commonly found in BU lesions . The extent of secondary infections in BU and their contribution to frequently observed delays in healing has not been studied so far . Here we have analyzed BU lesions before , during and after antimicrobial treatment for the presence of secondary infection .
The participants involved in the study were recruited from the Amasaman District Hospital and the Obom Health Centre in the Ga-West and Ga South Municipality , respectively . The participants were all laboratory confirmed BU cases and the analyzed samples fall into three main categories: 1 ) samples from 53 BU patients recruited consecutively before treatment; 2 ) samples from 20 BU patients recruited consecutively between four and six weeks after start of SR8 and 3 ) samples from 31 BU patients whose lesions were clinically suspected of secondary infection after SR8 treatment . Some of the participants overlapped in some of the categories: 71 of the participants were sampled once for analysis , 12 twice and 3 thrice within the study period , thus in total 104 individual samples , 84 swabs and 20 tissue samples , from 86 participants were analyzed . The swabs were obtained from 52 cases before treatment , 20 cases during and 12 cases after treatment and analyzed microbiologically ( Table S1 ) . The tissue samples for histopathological analysis were obtained from one case before treatment and 20 cases after treatment . Except for one sample taken after treatment , all tissue samples were also analyzed microbiologically ( Table S1 ) . A detailed questionnaire was used to obtain standard demographic data , document the clinical presentation of lesions and other lesion characteristics . Altogether the study involved 86 participants comprising 32 ( 37% ) females and 54 ( 63% ) males . The females' age ranged between two and 72 years and the males were between four months and 82 years . Median age for both groups was 33 years . Seventy-seven of the cases had lesions located on the limbs , three in the head and neck region , and one each located on the buttocks , armpit and back respectively; the lesion location of three participants was not documented . Only 2/86 patients were pre-ulcerative . These lesions , one nodule and one plaque , were sampled later during surgery . The remaining 84 patients had ulcers; 78 of them had only ulcers , one had an ulcer and a nodule , three had ulcers with edema , and two had ulcers with osteomyelitis . Based on the judgment of the responsible clinician , surgical debridement was performed for 1 patient prior to treatment and for 20 patients after completion of SR8 . Biopsy samples were collected in each instance for histopathological analysis ( Figure 1 ) . Laboratory confirmation of BU disease was done by IS2404 PCR and Ziehl-Neelsen microscopy as previously described [20] , [21] . Three swab samples were collected from clinically suspected ulcerative cases before treatment; one for IS2404-PCR based confirmation of BU , one for preparation of a direct smear for microscopic examination for the detection of bacteria and neutrophils after Gram staining ( Figure S1 ) , and the third was inserted into a sterile tube containing 3 ml of PBS for enumeration of the bacterial burden and the isolation of specific bacterial species . All swab specimens were collected from the undermined edges of lesions by first moistening the swab with sterile PBS using the Levine method of collecting swab specimen [22] . This has been found to be the best method for taking swabs as it is more reflective of tissue bio-burden as compared to other methods [23] . After cleaning the wound surface with normal saline , a swab was rotated over a 1 cm2 area with sufficient pressure to collect the fluid from within the wound tissue . From cases that were sampled during treatment and those that were clinically suspected of having a bacterial infection after completion of SR8 , three swab specimens were collected before surgery , and treated as above , except for the procedures for the laboratory confirmation of BU disease by PCR , since all cases had been previously confirmed as BU within the framework of a bigger study . From SR8 treated patients that underwent surgical management , tissue sample were analyzed if there was clinical suspicion of a secondary bacterial infection . While one sample was aseptically transferred into a clean sterile tube for enumeration of the bacterial load and species identification , a second sample was directly transferred into 10% neutral buffered formalin for histopathological analysis . The samples for bacteriological analysis were placed in an ice chest with ice packs to prevent bacterial multiplication and transported to the Bacteriology Department of the Noguchi Memorial Institute for Medical Research ( NMIMR ) for analysis , Tissue samples for were shipped to the Swiss Tropical and Public Health Institute for histopathological analysis . Ethical clearance was obtained from the institutional review board of the Noguchi Memorial Institute for Medical Research ( Federal-wide Assurance number FWA00001824 ) . The procedures for sampling in this study were essentially the same as those used in routine management of BU in Ghana . However , written informed consent was collected from all participants before study inclusion . In the case of children below sixteen years , written informed consent was collected from their parents or guardians . Patients were assured of the confidentiality of all information collected during the study . When swab samples reached the microbiology laboratory , the volume of PBS was topped up to 5 ml and both the swab and the PBS were transferred into a sterile glass tissue culture tube containing glass beads . The tubes were vortexed for about two minutes to dislodge any particles that were sticking to the swabs . Using the resulting stock suspension , serial dilutions from 10−2 to 10−6 were prepared . Hundred microlitres of serial dilutions of the swab or tissue suspensions were transferred into sterile Petri dishes and inoculated by the pour plate method using Plate Count Agar for total aerobic counts . The agar was left on the lab bench to set after which it was incubated at 37°C for 18–24 hours . The remaining 10−1 dilution of the suspension was centrifuged at 8 , 000 g for 25 minutes and after decanting , the pellet was inoculated onto MacConkey , Blood and Chocolate agar and incubated under aerobic conditions . The aerobic agar plates were examined after 24 hours and growing colonies were subcultured on Blood and MacConkey agar plates to obtain pure cultures . After incubation , the plates were examined using a colony counting chamber ( Gallenkamp , UK ) and those with colony counts between 30 and 300 were selected for computing CFU/ml or CFU/g , respectively , by multiplying the counts by the dilution factors . The lesion from which the sample was taken was classified as clean , contaminated or infected as indicated in the data analysis section . For tissue specimen , one gram of sample was weighed in a sterile plastic stomacher bag . Nine milliliters of PBS were added , samples were macerated in a stomacher and the resulting suspension was transferred into a sterile test-tube . Using this stock suspension , serial dilutions were prepared and plated out . Distinct bacterial colonies from the Blood and MacConkey agar plates were purified on Nutrient agar plates for identification . Bacterial isolates were Gram stained [24] and identified by biochemical tests as well as by molecular methods . Gram negative rod isolates were characterized by cytochrome oxidase analysis , and with Analytical Profile Index ( API 20E ) strips ( bio-Mérieux SA , Marcy-l'E'toile , France ) according to the manufacturer's instructions . Gram positive cocci were analyzed after Gram staining using the catalase test to differentiate between Staphylococcus spp . and Streptococcus spp . In order to further discriminate the catalase positive Gram positive cocci and especially to identify Staphylococcus spp . , the Staphylase kit Prolex Latex Agglutination System ( Pro-Lab Diagnostics ) was used . Gram positive bacteria were further characterized using the Hain Lifescience Genotype Product series for Gram positive bacteria Genotype BC Gram positive version 3 . 0 and Genotype staphylococcus version 2 test kits ( Hain Lifescience , Germany ) . Where species identification failed with the analytical profile index and the other biochemical assays , identification was achieved by MALDI-TOF mass spectrometry [25] . Susceptibility of isolates to specific drugs was tested using the Kirby-Bauer disc diffusion method on Mueller Hinton agar [26] . Sensitivity was tested against antibiotics such as Cotrimoxazole , Ampicillin , Tetracycline , Ciprofloxacin , Amikacin , Gentamicin , Penicillin , Erythromycin , Cefuroxime , Cefixime , Ceftriaxone , Chloramphenicol and Flucloxacillin . In addition Gram positive cocci were tested against methicillin and vancomycin . The results of isolation and drug sensitivity tests were provided to the treating clinician at the collaborating health facility . Since the locally available disc systems varied in coverage , some antibiotics were only tested with a subset of isolates . One limitation of this study is that we did not test for susceptibility against streptomycin and rifampicin . Histopathological analysis was done for all SR8 treated patients needing surgical management and presenting with a lesion clinically suspicious for secondary infection . Surgically excised tissue samples were immediately fixed after excision in 10% neutral-buffered formalin for 24 h at room temperature to maintain tissue structures . Afterwards samples were directly transferred to 70% ethanol for storage and transport . Tissue specimens were subsequently dehydrated , embedded into paraffin , and cut into 5 µm sections . After deparaffinization and rehydration , sections were stained with Ziehl-Neelsen/Methyleneblue ( ZN ) according to WHO standard protocols [3] . In this staining AFB appear pink and other bacteria are stained blue . Tissue sections were analyzed with a Leica DM2500 Microscope and pictures were either taken with a Leica DFC 420C camera or with an Aperio ScanScope XT . Recycled bandages from fifteen confirmed BU cases were collected conveniently before wound dressing for microbiological analysis . Ten grams bandage was weighed , added to 90 ml of sterile PBS and macerated with a laboratory blender to give a 10−1 dilution . Using this suspension , serial dilutions from 10−2 to 10−6 were prepared . Hundred microlitres of these serially diluted suspensions were transferred into sterile Petri dishes and inoculated by the pour plate method using Plate Count Agar for total aerobic counts . Bacterial enumerations were performed as described above . In addition the left over suspension was centrifuged at 3 , 000 g for 20 mins and the resulting pellet was plated for bacterial isolation . The values obtained from plate counts were computed into CFU/ml for wound exudates ( swabs ) or CFU/g for tissue sample . The antibiogram of each isolate was interpreted according to the manufacturer's specification as resistant , intermediate or susceptible . The percentages of cases in each category were then computed . Lesions were classified microbiologically as clean if no bacteria were isolated , as contaminated if bacterial counts were <106 CFU/g or ml and as infected if counts were >106 CFU/g or ml of specimen . Lesions were clinically classified as infected based on the following criteria: 1 . friable , bleeding granulation tissue despite appropriate care and management; 2 . purulent discharge ( yellow or green ) from wound or drain placed in wound; 3 . pain or tenderness , localized swelling ( edema ) , or redness/heat; 4 . tissue necrosis; 5 . skin grafting failure; abnormal odor coming from the wound site; delayed healing not previously anticipated . Twenty-four of the patients clinically classified as infected were in-patients and seven were out-patients , who were reporting twice a week for wound dressing . During wound dressing , the wounds were cleaned with normal saline to wash away debris . Wounds that appeared necrotic or had an offensive odor were cleaned again with vinegar and dressed with povidine iodine .
Swab samples of 52 consecutively recruited IS2404 PCR confirmed BU cases with ulcerative lesions were sampled before the commencement of SR8 treatment . Samples from three participants ( 5 . 7% ) did not yield any aerobic growth on plate count agar ( Table 1 ) . Seventeen ( 32 . 1% ) of the lesions with total CFU counts of 1 . 7×103 to 9 . 0×105 CFU/ml ( average 3 . 2×105 CFU/ml ) were microbiologically classified as contaminated . Microbiologically Infected lesions were observed in 33/52 patients ( 63 . 5% ) ; aerobic counts from this group ranged between 1 . 0×106 to 3 . 5×109 CFU/ml with an average value of 1 . 1×109 CFU/ml . The most frequently identified bacterial species from the infected lesions prior to start of treatment ( Table 1 ) were S . aureus ( n = 9; 21 . 4% ) , P . aeruginosa ( n = 7; 16 . 7% ) and P . mirabilis ( n = 6; 14 . 3% ) . The responsible clinician decided to perform wound debridement of one of the lesions prior to SR8 initiation , since it showed clinical signs of a strong secondary infection ( Figure 1D ) . A biopsy specimen was taken and the histopathological analysis of the tissue sample ( Figure 1A–C ) revealed , typical hallmarks of BU , such as fat cell ghosts , tissue necrosis and epidermal hyperplasia ( Figure 1A ) . In addition , clusters of cocci were observed in the subcutaneous tissue between the fat cells ( Figure 1A box , B , C ) . This area probably represents the tissue base of the undermined edges . These findings correlated well with the microbiological analysis , since S . aureus was isolated in large numbers from the lesion ( 1 . 2×109 CFU/g ) . Twenty laboratory-confirmed BU cases were consecutively sampled between four and six weeks after start of SR8 treatment and analyzed for infection of the lesions . Of these lesions , 7/20 ( 35 . 0% ) and 13/20 ( 65 . 0% ) were microbiologically classified as contaminated or infected , respectively; clean wounds were not observed ( Table 1 ) . The aerobic bacterial load ranged between 1 . 5×106 and 3 . 5×109 CFU/ml , with an average value of 5 . 6×108 CFU/ml for the microbiologically infected lesions . The contaminated lesions had counts between 5 . 2×103 and 7 . 3×105 CFU/ml ( average 3 . 3×105 CFU/ml ) . Also here P . aeruginosa ( n = 6; 35 . 3% ) and P . mirabilis ( n = 2; 11 . 8% ) , but not S . aureus ( n = 0 ) , were the most frequently identified bacterial species isolated from the infected lesions ( Table 1 ) . Thirty-one BU lesions with clinical signs of secondary bacterial infection after completion of SR8 treatment were sampled for laboratory investigation . Clinical signs indicative for secondary infection were documented for 28 of them and included: localized pain ( 28/28 ) , viscous/purulent discharge ( 28/28 ) , edema ( 5/28 ) and localized heat ( 4/28 ) . In addition , delayed healing not previously anticipated ( 17/28 ) , offensive odor ( 15/28 ) and discoloration of tissues both within and at the wound margins ( 3/28 ) were regarded as signs of secondary infection ( Table 2 ) . The time at which infection was detected ranged from a few weeks to fifteen months after completion of SR8 . Seven ( 22 . 6% ) of the 31 lesions clinically suspected to be infected were not confirmed microbiologically by aerobic bacterial count analysis , as the total plate count ranged only between 1 . 3×103 and 8 . 9×105 CFU/ml ( average 2 . 7×105 CFU/ml ) . The remaining twenty-four ( 77 . 4% ) lesions that were microbiologically confirmed as infected had plate counts ranging between 1 . 2×106 and 3 . 5×109 CFU/ml ( average value of 1 . 2×109 ) . P . aeruginosa ( n = 8; 32% ) , P . mirabilis ( n = 5; 20% ) and S . aureus ( n = 3; 12% ) dominated among the isolates . The bacterial load observed in cases analyzed within four weeks post SR8 ranged between 1 . 3×103 and 4 . 0×109 CFU/ml; that between five and 12 weeks was between 9 . 3×104 and 1 . 2×109 CFU/ml; and that between 9 and 15 months post SR8 ranged between 2 . 7×106 and 1 . 8×109 CFU/ml . Nineteen tissue samples and 12 swab samples were analyzed ( Table S1 ) and the bacterial load ranged between 1 . 3×103 and 4 . 0×109 CFU/ml for tissues and between 5 . 2×107 and 2 . 1×109 for swabs . Tissue samples from 20/31 of the microbiologically analyzed lesions showing clinical signs of secondary infection after completion of SR8 were also analyzed by histopathology , since the responsible clinicians decided to perform a wound debridement . Microbiological analysis had categorized 16 of these lesions as infected and four as contaminated . None of the microbiologically contaminated wounds presented in the histopathological analysis with a detectable secondary infection . In contrast 12/16 ( 75% ) of the lesions classified microbiologically as infected presented with an infection either with cocci , rods or both ( Table 2 ) . Infection was mainly observed in the stratum corneum ( 6/12; 50% ) or on the open ulcer surface ( 3/12; 25% ) and only rarely ( 3/12; 25% ) deeper inside the excised tissue ( Table 2 ) . Histopathological analysis of specimen from patient 9 ( Figure 2 A–D ) revealed a layer of densely packed rods at the open ulcer surface visible already at low magnification as an intensely blue stained band ( Figure 2B ) At higher magnification , clusters of rod shaped bacteria were observed ( Figure 2 C , D ) . Microbiological analysis confirmed the presence of P . aeruginosa . Tissue excised from patient 16 ( Figure 2E–H ) showed a double infection: cocci being present inside the stratum corneum ( data not shown ) as well as an extensive infection of the dermal and subcutaneous tissue with rods ( Figure F–H ) . Microbiological analysis isolated S . aureus as well as Gram-negative rods . In most of our analysis , histopathological and microbiological results showed a good correlation for most of the patients ( Table 2 ) . Using the disc diffusion assay , a total of 98 Gram-negative rods and Gram-positive cocci obtained from BU wounds were tested for resistance against antibiotics commonly used in Ghana . None of the isolates tested was sensitive to all drugs included in the analysis ( Table 3 ) . Five Gram-negative rods were resistant to all tested drugs . More than 70% of the 18 S . aureus isolates obtained from infected ( n = 12 ) or contaminated ( n = 6 ) lesions were resistant to flucoxacillin , ampicillin and penicillin . In contrast , 15/18 ( 83% ) were susceptible to gentamicin . The prevalence of S aureus isolates resistant to methicillin ( MRSA ) and vancomycin ( VRSA ) was 33% and 17% , respectively . Likewise most of the P . aeruginosa strains were resistant to most of the tested drugs . However , most isolates ( 18/22; 82% ) were susceptible to gentamicin . Results for the other Gram-negative and -positive bacteria are provided in Table 3 . When monitoring wound management procedures , it was realized that patients and care-givers were instructed by health workers to wash and recycle dressing bandages . We therefore conveniently sampled dressings that have been used and washed for the next dressing . Seventeen bandages from fifteen BU cases were analyzed and as shown in Table 4 , all of them had some bacterial contamination with total aerobic plate counts ranging between 2 . 2×103 and 3 . 2×108 CFU/g with an average count of 2 . 8×107 and a median value of 1 . 2×105 CFU/g . While bacterial species identified included commensals such as staphylase negative Staphylococcus spp . , also potential pathogens including S . aureus , P . aeruginosa , Flavibacterium oryzihabitans , Enterobacter agglomerans and Enterobacter cloacae were isolated . The drug susceptibility patterns of isolates are indicated in Table 4 . Similar isolates were also isolated from patients' wounds .
Mycolactone , the cytotoxic macrolide toxin of M . ulcerans plays a key role in the pathology of BU . It causes apoptosis of mammalian cells [8] , [27] and has immunomodulatory activity [28] , [29] . Since a number of macrolides have antibiotic activity against a broad spectrum of bacteria , including streptococci , pneumococci , staphylococci , enterococci , mycoplasma , mycobacteria , rickettsia , and chlamydia [30] , it has been speculated that mycolactone secreted by M . ulcerans during active disease prevents secondary bacterial infections of BU lesions . The goal of this study was to find out whether ulcerative BU lesions are indeed rarely colonized or infected by other bacterial species . To address this , BU wounds were characterized before SR8 treatment by both direct smear microscopic analysis for the presence of bacteria and neutrophils [20] and by pour plate determination of aerobic CFU counts . More than 60% of the lesions tested before treatment had bacterial counts ≥106 CFU/ml and direct smear examination frequently showed large numbers of bacteria and neutrophils ( Figure S1 ) . A broad spectrum of bacterial species was isolated from the lesions with S . aureus , P . aeruginosa and P . mirabilis being the most frequently found species . This suggests that M . ulcerans infection and mycolactone secretion does not prevent secondary bacterial infections . Chronic wounds often have a bacterial burden that is massively exceeding levels used to define lower limits for the definition of infection in acute surgical wounds ( i . e . 106 CFU/g of tissue ) . However , many chronic wounds go on to closure despite levels of infecting microorganisms ≥108 CFU/g of tissue , with infection by Group B streptococci being one exception to this rule [12] , [13] , [16] . Because of the intrinsic differences in the way acute and chronic wounds respond to the burden of microorganism , emphasis is currently being placed on holistic assessments , with clinical signs and symptoms playing key roles in the diagnosis of chronic wound infection . Clinical signs usually employed for diagnosis include erythema , edema , heat , purulent exudates with concurrent inflammation , pain , delayed healing , discoloration of granulation tissue , friable granulation tissue , pocketing at the base of the wound , foul odor , and wound breakdown [13] , [14] , [17] . In particular increasing pain and wound breakdown have been shown to be good predictors of infection in chronic wounds . In this study we combined clinical , histopathological , qualitative and quantitative microbiological methods to analyze BU lesions for the presence of infections after completion of SR8 treatment . Lesions from 28 patients showing clinical signs of infection were included in this analysis . 75% of these lesions yielded CFU counts >106 CFU/ml ( average value of 1 . 2×109 ) and frequently species with pathogenic potential , such as S . aureus , P . aeruginosa , S . haemolyticus , E . cloacae and K . pneumonia were isolated . Pain and yellow discharge turned out to be highly predictive clinical indicators for infection . For the patients that had clinical signs of infection after SR8 , culture and drug susceptibility testing results were submitted to the treating officer . However documentation of the treatment and subsequent follow-up of patients was beyond the scope of this study . A study analyzing the microbial flora of healing and non-healing decubitus ulcers [31] found S . aureus , Streptococcus spp . , E . coli , Klebsiella spp . , Proteus spp . and P . aeruginosa as the main organisms that caused infection of the ulcers . Chronic venous ulcers have been found to be infected with S . aureus , P . aeruginosa , Coagulase-negative staphylococci , Proteus spp . and anaerobic bacteria [32] . Thus most of the organisms isolated in this study from BU lesions have also been found associated with infection of other types of wounds . Similar to what has been reported in other studies [33] , lesions were in many cases infected with more than one bacterial species ( Table 2 ) . Our data on the microflora of lesions upon admission indicate that BU lesions may be contaminated from the communities as a result of improper wound care practices by the patients in their quest to treat the infection either on their own or with the help of traditional healers or herbalists . There is major concern about subsequent acquisition of antibiotic resistant organisms from the hospital settings . After the present pilot study demonstrating colonization and infection during and after SR8 treatment , we plan to perform longitudinal studies with patient cohorts to study the influence of BU wound management practices on secondary bacterial infections . The method used for collecting wound specimens can influence the data obtained from microbiological culturing . Currently , collection of a biopsy specimen is the gold standard for determining the presence and identity of microorganisms within the wound bed tissue [12] , [16] , [34]–[37] . However , there are limitations as to which healthcare providers can collect biopsies , the availability of laboratories offering microbiological culture testing on biopsies , the expenses involved with the performance of these tests , and the potential for further tissue damage and delay of wound healing when biopsies are taken . In the present study we employed swabbing [22] , [36] as the main sampling procedure and performed histopathological studies with tissue specimen only from 20 cases that underwent surgical intervention . The histopathological analysis detected bacterial populations in 75% ( 12/16 ) of the analyzed lesions classified as infected and in none ( 0/4 ) of the lesions classified as contaminated . This strong correlation between results obtained with tissue and swab samples confirms results of previous studies [23] indicating that microbiological swabbing is a good sampling procedure for the determination of infection of wounds . Histopathological analysis detected infecting bacteria populations only rarely deeper inside the excised tissue and mainly in the stratum corneum or on the open ulcer surface , where bacteria are accessible for the swabs . Contamination of BU lesions prior to SR8 treatment may be a result of wound care practices by the patients . Also during SR8 treatment a range of bacterial species , with Gram-negative rods dominating , were isolated from the lesions . This indicates that SR8 does not necessarily eliminate contamination or secondary infection of lesions . Bacterial species , such as P . aeruginosa , K . pneumoniae and S . aureus isolated from infected lesions after completion of SR8 treatment , may however also have been acquired from the hospital setting . A detailed characterization of isolates is required to address this important issue further . Both mono and multiple antibiotic resistant strains were isolated with high frequency from the BU lesions . For example all the tested S . aureus strains were resistant to penicillin , 22% were methicillin resistant and 17% vancomycin resistant . Dependent on the setting , both lower ( Nigeria , [38] ) or higher ( South-Africa , [39] ) frequencies have been reported in Africa . Most worrying in this context is the high ( 83% ) level of resistance of S . aureus isolates to flucloxacillin , which is in Ghana the main antibiotic in use for treating skin infections such as boils and cellulitis . In addition , we acknowledge that true VRSA is rare , and that the occurrence of apparent VRSA is being followed up through referral of isolates to an international reference laboratory . Postoperative infections of wounds represent the commonest surgical complication causing substantial increases in the duration and costs of hospital stays [40] . Our pilot study involving BU patients at different time points of SR8 treatment indicates that secondary bacterial infection may be a prominent cause for delays in wound healing and skin grafting failures . These findings call for an optimization of BU wound management and hygiene procedures to better control secondary infections . Also the choice of treatment of secondary infections with locally available antimicrobial agents requires a better understanding of the infecting flora and of drug susceptibility patterns . Our study did not follow the same patients from beginning of treatment till they were healed and this has limited the ability to determine causes and consequences of wound infection . More studies are required to ascertain the impact and source of wound infection in SR8 treatment of BU and to support development of guidelines for wound care in BU case management . In addition to wounds we also analyzed bandages that have been washed by the patients themselves to be re-used for wound dressing . From these bandages we isolated potential wound pathogens including S . aureus , P . aeruginosa , Flavibacterium oryzihabitans , Enterobacter agglomerans and Enterobacter cloaca; thus the bacteria profile of the wound samples was comparable to that of the bandages . These findings indicate that the recycling of bandages may not be a good practice as it may be one of the sources of wound infection . We recommend that if for economical reasons bandages need to be recycled , they must be washed well with an appropriate disinfectant . | Buruli ulcer ( BU ) can lead to large ulcerative lesions due to extensive skin loss caused by the necrotizing effect of the main virulence factor mycolactone . For a long time the general perception was that BU lesions are not infected by other bacteria because of a postulated antimicrobial effect of the macrolide toxin , mycolactone . In this study , we analyzed laboratory confirmed BU lesions before , during , and after streptomycin/rifampicin treatment . Contrary to popular belief , our findings show that BU lesions are frequently co-colonized with other potential bacterial pathogens before , during , and after antibiotic treatment . For example , 75% of cases that were clinically indicative of being infected after treatment were microbiologically confirmed as infected . Most microbiologically infected cases were also confirmed by histopathological analysis . The most prominent bacterial species isolated included Pseudomonas aeruginosa , S . aureus , and P . mirabilis . When we tested the isolates against first line drugs used in Ghana , the isolates were found to be resistant to most of these drugs . This study indicates that wound care practices need to be improved and that wound infection may be a common cause of wound healing delay in BU . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [
"bacteriology",
"medicine",
"dermatology",
"infectious",
"diseases",
"buruli",
"ulcer",
"emerging",
"infectious",
"diseases",
"medical",
"microbiology",
"neglected",
"tropical",
"diseases",
"biology",
"microbiology"
] | 2013 | Secondary Bacterial Infections of Buruli Ulcer Lesions Before and After Chemotherapy with Streptomycin and Rifampicin |
Latent HIV infection of long-lived cells is a barrier to viral clearance . Hematopoietic stem and progenitor cells are a heterogeneous population of cells , some of which are long-lived . CXCR4-tropic HIVs infect a broad range of HSPC subtypes , including hematopoietic stem cells , which are multi-potent and long-lived . However , CCR5-tropic HIV infection is limited to more differentiated progenitor cells with life spans that are less well understood . Consistent with emerging data that restricted progenitor cells can be long-lived , we detected persistent HIV in restricted HSPC populations from optimally treated people . Further , genotypic and phenotypic analysis of amplified env alleles from donor samples indicated that both CXCR4- and CCR5-tropic viruses persisted in HSPCs . RNA profiling confirmed expression of HIV receptor RNA in a pattern that was consistent with in vitro and in vivo results . In addition , we characterized a CD4high HSPC sub-population that was preferentially targeted by a variety of CXCR4- and CCR5-tropic HIVs in vitro . Finally , we present strong evidence that HIV proviral genomes of both tropisms can be transmitted to CD4-negative daughter cells of multiple lineages in vivo . In some cases , the transmitted proviral genomes contained signature deletions that inactivated the virus , eliminating the possibility that coincidental infection explains the results . These data support a model in which both stem and non-stem cell progenitors serve as persistent reservoirs for CXCR4- and CCR5-tropic HIV proviral genomes that can be passed to daughter cells .
Long term combination anti-retroviral therapy ( cART ) blocks viral spread in vivo but is not curative , as plasma virus rebounds after cART interruption . Sequence analysis of residual circulating and rebounding virus in HIV+ patients indicates that virions likely come from the activation of latent provirus that had been archived since before the initiation of therapy rather than from low-level replication and spread of cART-resistant virus [1 , 2] . HIV enters cells via HIV Env interacting with CD4 plus a co-receptor , usually CCR5 or CXCR4 . CXCR4-utilizing viruses differ from those that utilize CCR5 in their ability to infect stem cells that can engraft and generate multiple lineages in a mouse xenograft model [3] . In contrast , CCR5-tropic viruses infect HSPCs that are restricted in their capacity to differentiate [3] . Recently , Nixon and colleagues elegantly demonstrated that myeloid progenitors , including common myeloid progenitors ( CMPs ) and granulocyte/monocyte progenitors ( GMPs ) , express CCR5 and can be infected by CCR5-tropic HIV in vitro and in a humanized mouse model [4] . Based largely on patterns of hematopoiesis that occur following transplantation , hematopoietic progenitors , such as those targeted by CCR5-tropic HIVs , were thought to be short-lived in vivo [3–5] . However , in situ tagging experiments in mice have recently found that non-stem cell progenitors make an enduring contribution to native hematopoiesis in adults through successive recruitment of thousands of clones , each with a minimal contribution to mature progeny [6–8] . Consistent with this , non-stem cell myeloid progenitors such as GMPs were found to persist in people with aplastic anemia despite dramatic losses of stem cells [6] . Thus , a large number of long-lived progenitors , rather than classically defined Hematopoietic stem cells ( HSCs ) , may be the main drivers of steady-state hematopoiesis during adulthood [7 , 8] . Here , we provide evidence that non-stem cell hematopoietic progenitors harbor CCR5-tropic HIVs for years in optimally treated people , providing new evidence that non-stem cell progenitors are long-lived in people without evidence of bone marrow disease and can potentially serve as reservoirs of HIV . We also demonstrate that CD4high HSPC subsets that we show include multi-potent progenitors ( MPPs ) are preferentially targeted by both HIV subtypes in vitro . Moreover , we provide in vivo evidence that infected HSPCs can differentiate into multiple lineages that harbor provirus . These data expand our understanding of HIV infection and hematopoiesis by demonstrating that in addition to stem cells , intermediate progenitor cells potentially provide an enduring reservoir for CCR5- and CXCR4-tropic HIV proviral genomes .
To better understand the types of hematopoietic stem and progenitor cells ( HSPCs ) that are infected by HIV in vivo , we developed an approach to efficiently isolate HSPC populations enriched ( Sort 1 ) or depleted ( Sort 2 ) for stem cells ( Fig 1A and 1B ) . Compared to Sort 1 cells , Sort 2 cells expressed lower levels of CD133 ( Fig 1B ) and were depleted for hematopoietic stem cells ( HSCs ) and multi-potent progenitors ( HSC/MPPs ) ( Fig 1C–1G ) . Conversely , Sort 2 cells were enriched for more restricted progenitors ( common myeloid progenitors ( CMPs ) and megakaryocyte/erythrocyte progenitors ( MEPs ) ( Fig 1H and 1I ) [9] . Enrichment of MEPs in Sort 2 samples was confirmed using methylcellulose colony formation assays ( Fig 1J ) . To develop a better understanding of which HSPCs harbor HIV in vivo , we obtained samples from 47 HIV-infected donors , including two that had been initially treated during acute infection . All donors were on therapy with undetectable viral loads for least six months . A 20 ml bone marrow sample and 100 ml of peripheral blood were collected from each donor . HSPCs were isolated from adherence depleted bone marrow mononuclear cells in two steps as described in Fig 1 . From 20 cc of bone marrow , we obtained ~2 . 5x106 total HSPCs per donor . For 41 of 47 donors , we obtained adequate aspirates and the purified HSPCs met our criteria of having <1% CD3+ T cell contamination and >80% CD34+ or CD133+ cells . The mean purity of included samples was approximately 94% CD133 for Sort 1 and 90% CD34 for Sort 2 ( Table 1 , S1 and S2 Tables ) . DNA was isolated from each sample and multiplex single genome amplification ( SGA ) polymerase chain reaction ( PCR ) was used to amplify gag and env amplicons or near full-length genomes . For each donor , we selected a primer pair combination that most efficiently amplified HIV sequences from peripheral blood mononuclear cell ( PBMC ) DNA prior to testing HSPC samples . After analyzing at least 80 , 000 cells from all samples that met our purity criteria , we determined that most donors ( n = 24 , 59% ) had detectable HIV provirus in HSPCs . More cells were screened in the positive group than in the undetectable group ( 975 , 959 versus 661 , 965 ) but that difference and the level of sample purity between the two groups were not statistically significant ( Table 1 ) . Further , the timing of HAART was not a significant factor in our ability to detect provirus; one of two donors treated since acute infection had detectable HSPC-associated provirus and provirus was present in long term suppressed patients ( up to 9 . 8 years , Table 1 ) . The overall mean frequency of provirus in HSPCs was 2 . 4 copies per million cells based on the number of positive 1st round PCR reactions ( 82 ) set up at limiting dilution that produced a gag and/or env amplicon out of the total number of cells assayed ( 35 million ) . For individual donors , the frequency ranged from <1 per 1 . 3 x106 cells to 18 copies per 106 cells . To rule out T cells as a source of HIV DNA in HSPC samples , we eliminated all HSPC samples with >1% contaminating CD3+ cells and all samples included in our final analysis contained <0 . 52% CD3+ cells ( S1 and S2 Tables ) . In addition , we used a previously published statistical method that takes into account HIV genome frequency in sorted and flow-through samples , assigning a p value to indicate the likelihood that HIV DNA in HSPC samples came from T cells [10] . This analysis is shown in detail in Fig 2 . Briefly , we carefully assessed the frequency of CD3+ T cells and provirus in both the sorted sample and in the flow-through sample . Then , we compared the frequencies assuming that only CD3+ T cells account for all provirus . As shown in Fig 2B , the frequency of infected CD3+ T cells would have to have been much higher in the sorted sample than in the flow-through to account for the provirus in the sorted HSPC samples ( e . g . 1 in 52 versus 1 in 15 , 000 for donor 409000 ) . This difference is assigned a p value that takes into account 95% confidence intervals and only samples with p<0 . 05 were included in our final analysis . Consistent with our conclusion that HIV DNA from Sort 1 and 2 came from HSPCs and not CD3+ T cells , we observed no correlation between proviral frequency in the samples and the frequency of contaminating CD3+ cells ( S2 Fig ) . In addition , rearranged T cell receptor PCR assays were performed to confirm that near-full-length genomes from HSPC DNA samples were unlikely to have originated from T cells ( Fig 2C ) . Similar results were obtained from donor 413402 , which was screened by PCR because too few cells were available to accurately assess this sample by flow cytometry ( S4 Table , S3 Fig ) . [The caveats for the rearranged TCR PCR assay are that it is not quantitative and it is associated with non-specific background bands that limit the amount of DNA that can be added to the reaction . These non-specific bands arise in all samples , including negative control HEK 293 cells , and are not related to TCR based on sequencing analysis . Because of these limitations , the statistical analysis we described in Fig 2 provides a more robust and quantitative assessment . ] Based on initial results that CCR5-tropic HIVs infect non-stem-cell progenitors that were originally believed to be short lived , we expected to mainly observe CXCR4-tropic virus in HSPC preparations . To assess this , we examined env amplicons available to study from a subset of 19 donors from the overall cohort . As summarized in Tables 2 and 3 , we isolated a total of 52 env C2-V3 amplicons . Each amplicon was assigned a genotype using the indicated co-receptor prediction software ( Table 3 ) . 16 amplicons from 8 donors were predicted to be CXCR4-tropic , including three near-full-length genomes with full open reading frames and cis elements ( S3 Table ) . Unexpectedly , we also isolated a total of 36 amplicons from 17 donors that were predicted to be CCR5-tropic , including one near-full-length genome with full open reading frames and cis elements ( S3 Table ) . Overall , the genotopyes of env amplicons from HSPCs closely matched those from peripheral blood mononuclear cells for each donor ( Table 3 ) . Because env genotype prediction tools are not always reliable , we confirmed Env tropism with a phenotypic assay . For this analysis , we used either HSPC-derived full-length Env or a non-HSPC-derived Env with identical nucleotide or amino acid V3 region from the same donor as available ( Table 4 ) . A phenotypic assay utilizing 3T3 cells expressing CD4 and individual chemokine receptors [11] was used for this assessment . This assay confirmed the tropism of ten CCR5-tropic Envs , four CXCR4/dual tropic Envs and demonstrated that one Env was not functional . The isolation of HIV encoding Envs of both tropisms from HSPCs suggests either that CCR5-tropic Envs unexpectedly target HSCs or that restricted progenitor cells targeted by CCR5-tropic viruses survive longer in vivo than expected . To better understand whether CCR5-tropic viruses might target restricted progenitors that persist longer than expected , we asked whether provirus could be detected in Sort 2 , which contained restricted progenitors that were unlikely to be stem cells . Interestingly , we found no significant difference in the number of donors with amplicons in Sort 1 versus Sort 2 subsets [14 donors had amplicons isolated from Sort 1 and 11 had amplicons isolated from Sort 2 , ( Tables 2 and 3 ) ] . The mean frequency of env amplicons was higher in Sort 1 than Sort 2 but this difference did not achieve significance . ( The mean frequency was four copies per million cells for Sort 1 versus two copies per million cells for Sort 2 , p = 0 . 06 . ) None of the amplicons isolated from Sort 2 were identical to those from Sort 1 , indicating that independent infection of restricted progenitors rather than differentiation of infected stem cells explains the presence of provirus in this population . In sum , these results suggest that non-stem cell restricted HSPCs can be infected by HIV and endure for at least the period of effective antiretroviral treatment . The result that HSPCs depleted of HSCs harbor HIV provirus that persists in optimally treated people as well as the finding that CCR5-utilizing virus persists in HSPCs was unexpected; therefore , we pursued additional evidence to better understand this finding . First , we assessed expression of HIV receptors in HSPC subsets . To accomplish this , we used a publicly available microarray dataset of RNA expression in human bone marrow HSPCs [12] and used established markers to purify murine bone marrow HSPCs for an RNA-seq analysis to profile expression of HIV receptors in HSPC subtypes . After confirming that progenitor subsets from each species expressed the expected developmentally appropriate set of genes ( Fig 3 ) we found that both approaches yielded similar results . As shown in Fig 3 , both revealed very low CCR5 expression in HSCs with higher expression in some restricted hematopoietic progenitor sub-types . These results agree with published studies showing low or no expression of CCR5 protein by HSC-enriched cells [3 , 13] with more expression of CCR5 protein in restricted hematopoietic progenitor sub-populations [4 , 13] . In addition , both approaches showed that CXCR4 and CD4 RNA were expressed by HSCs and several other progenitor populations ( Fig 3 ) . Based on this analysis , CXCR4-tropic viruses are predicted to target a wide range of progenitor subsets including HSCs whereas CCR5-tropic Envs are more likely to target restricted HSPC subsets such as GMPs . In prior studies , we used pseudotyped lentiviral reporter constructs to examine differential targeting of HSPCs by CCR5 and CXCR4 and it remained possible that full length , wild type HIVs target cells differently . To examine this question , we compared HIV infection of HSPCs by two wild type viruses , NL4-3 ( CXCR4-tropic ) and YU-2 ( CCR5-tropic ) ( Fig 4A ) . After demonstrating that CD133bright cell populations contain the majority of HSCs based on CD38 , CD45RA and CD90 staining ( Fig 4B ) , we used the level of CD133 staining to assess HIV infection of HSCs . As shown in Fig 4C , full length HIVs demonstrated the same pattern as previously observed using pseudotyped lentiviral vectors; CCR5-tropic YU2 infected a restricted pattern of progenitors depleted of stem cells whereas NL4-3 targeted a wide range of progenitors , including those likely to be stem cells . ( Maraviroc and AMD3100 appropriately inhibited entry via CCR5 and CXCR4 respectively , Fig 4C , lower panels . ) Correspondingly , on average , we measured about 4 . 5 times more CD133 on HSPCs infected by NL4-3 than those infected with YU2 ( Fig 4D and 4E ) . Further , the same pattern was observed using a lentiviral construct ( HIV-7/SF-GFP ) pseudotyped with additional Env proteins including one from a CCR5-tropic transmitted/founder virus [SVPB16 ( SV16 ) ][3 , 14 , 15 ( Fig 5 ) . In sum , consistent with prior results , CCR5-tropic viruses consistently demonstrated a restricted pattern of infection of more differentiated progenitors that contrasts with the wide range of progenitors targeted by CXCR4-tropic and VSV-G-pseudotyped viruses . Confirmation that a wide range of CCR5-tropic HIVs are restricted to non-stem cell HSPCs suggests that the CCR5-tropic HIV we detected in stem cell–depleted HSPC populations from patients likely came from more restricted progenitors that survived longer than previously appreciated and that these cells might also serve as long lived cellular reservoirs of HIV . The HIV receptor , CD4 , is usually required for infection and is expressed on CD34+ HSPCs , although at low levels compared to CD4+ T cells [16 , 17] . If the relative level of CD4 expression on HSPCs determined susceptibility of HSPCs to infection , then CD4 expression would serve as an indicator of the subtypes of HSPCs potentially targeted by HIV . To examine this question , we treated HSPCs with a GFP-expressing lentiviral vector pseudotyped with CCR5- or CXCR4-tropic Env proteins ( Fig 6A and 6B ) and assessed CD4 levels on the GFP+ transduced cells . We observed that HSPCs within a CD4high flow cytometric gate displayed 2–30 times greater infection than CD4low/- cells ( Fig 6A–6C ) . The increased infection of CD4high cells was not due to a greater capacity of these cells to support infection by this virus because the same virus pseudotyped with the vesicular stomatitis virus glycoprotein ( VSV-G ) demonstrated no such preference ( Fig 6B ) . Further , CCR5-tropic envelopes had a significantly greater propensity to target CD4high progenitors compared to CXCR4 and dual-tropic envelopes ( Fig 6B and 6C ) . Thus , relative CD4 expression levels correlated with susceptibility of HSPCs to infection by HIV-1 and HSPCs that express higher levels of CD4 are more likely to become infected . To determine whether CD4 marks a stable and separable progenitor subset with unique characteristics , we used flow cytometry to determine whether HSPCs could be separated into low and high CD4 expressing cells . Remarkably , sorting separated two distinct HSPC populations with different levels of CD4 ( Fig 7A ) . We then used these populations to demonstrate that CD4high HSPCs could form GEMM , granulocyte/macrophage ( GM ) , and erythroid ( E ) colonies ( Fig 7B ) . Thus , CD4 marks a subset of HSPCs that includes a number of different types of progenitors , including multipotent progenitors capable of generating multi-lineage GEMM colonies . To examine the CD4high sub-population in more detail , we used cell surface markers that had been validated with functional assays for HSPC subsets [9] . Remarkably , we found that CD4high HSPCs in Sort 1 contained a significantly greater frequency of HSCs and MPPs ( CD38-CD10-CD45RA- ) than CD4low HSPCs in the same Sort ( Fig 8 ) . Because CD133 also marks populations enriched for HSCs , we confirmed this result by demonstrating that there were significantly higher levels of CD4 on CD133high HSPCs than on CD133dim HSPCs ( ratio paired t test , p = 0 . 020 ) . In contrast , Sort 1 CD4low/- HSPCs and all Sort 2 cells that had lower levels of CD133 ( including those that were relatively CD4high ) were less frequently HSC/MPPs and more frequently restricted progenitors such as CMP/MEPs ( Fig 8 ) . Similar results were obtained whether or not lineage positive cells were depleted from the sample prior to analysis ( Fig 8C , open symbols ) . Thus , CD4 is expressed by a heterogeneous subset of hematopoietic progenitors and is expressed at significantly higher levels on subsets that include HSCs and MPPs . If HIV infects progenitor cells in vivo , HIV genomes could theoretically be passed to differentiated daughter cells as long as differentiation did not lead to reactivation of the virus from latency and cell death . To determine whether HIV can be transmitted by differentiation of infected progenitors , we assessed HIV proviral frequency in CD4-negative HSPC progeny . ( CD4-negative progeny were chosen for this analysis because cells lacking this HIV receptor are unlikely to be directly infected . ) To reduce the possibility of contamination by CD4-expressing cells , we depleted CD4+ cells using an anti-CD4 magnetic bead column prior to fluorescence activated cell sorting ( FACS ) . Following bead depletion and FACS , CD3+CD4+ T cells were undetectable in most samples ( Table 5 ) . Moreover , lineage-positive cells ( CD19+ B cells , CD8+ T cells and CD56+ natural killer ( NK ) cells ) were >98% CD4 negative ( indicated as “post-FACS” in Fig 9A ) . To determine whether HIV proviral DNA was present in these lineages , we used multiplex SGA PCR as described above . Remarkably , we generated a total of 38 LTR-gag or C2-V3env amplicons from four of five donors with CXCR4-tropic HIV but only one of five donors with only CCR5-tropic virus ( Table 5 ) . In two cases ( donors 420000 and 431000 ) , amplicons were identical to those isolated from HSPCs ( indicated as # in Table 5 ) . These cells were highly purified with undetectable CD3+CD4+ T cells ( Fig 9A and Table 5 ) . Using a quantitative statistical analysis , we found that the amplicons from CD4-negative lineages were unlikely to have come from contaminating CD3+CD4+ T cells ( p<0 . 05-p<0 . 001 , Table 5 ) . These results provide , strong evidence that HIV provirus can be transmitted from infected progenitors to progeny cells in vivo . Although we only detected provirus in CD4-negative cells from one of five donors ( 431000 ) with predominantly CCR5-tropic HIV , this donor provided the strongest evidence for HIV infection of multi-potent progenitors . Indeed , using SGA PCR , we amplified 14 identical CCR5-tropic C2-V3env amplicons from all three CD4-negative lineages , which were perfect matches to one another as well as to an amplicon isolated from HSPCs [Figs 9B and 10A] . In addition , seven first round SGA multiplex PCR reactions generated both C2-V3env amplicons as well as LTR-gag amplicons , all of which were identical ( Table 5 , Figs 9C and 10A ) . Remarkably , these amplicons contained a signature 469 bp deletion that removed the packaging site , the major splice site and the gag start codon , effectively inactivating the virus ( Fig 10A ) . We confirmed that the deleted gag came from the same proviral genome as the C2-V3env amplicons by using SGA PCR to isolate two near-full-length genome amplicons from CD4-negative cells ( Fig 10A ) . The presence of replication defective clonal proviral genomes in multiple differentiated hematopoietic lineages and in HSPCs provides strong evidence that infected multi-potent progenitors persist and differentiate in optimally treated people . A phylogenetic analysis of all donor sequences ensured that all donor 431 sequences clustered together , ruling out contamination and cross contamination as confounding factors ( S1 Fig ) . Moreover , phylogenetic analysis revealed that amplicons isolated from CD4-negative cells ( B , NK and CD8 ) were not common in CD4+ cells or unfractionated PBMCs , making cross-contamination an unlikely explanation for their relatively high frequency in CD4-negative lineages ( Fig 9B and 9C ) . Further , we used previously described statistical analysis [10] to demonstrate that the LTR-gag amplicons from B and NK cells were unlikely to have come from contaminating CD8+ T cells [p<0 . 05 ( 1 ) ] .
The identification and characterization of cell types harboring HIV genomes is crucial for the development of strategies to promote clearance . HSPCs support both active and latent infection by HIV in vitro and in vivo [13 , 18] . However , prior studies suggested a model in which only CXCR4-tropic viruses , which infect long-lived HSCs would be capable of persisting in vivo [3] . Here , we provide evidence that non-stem cell CD34+ progenitors infected by CCR5-tropic viruses are also long-lived . Indeed , HIV provirus isolated from HIV-infected people treated with cART for years was often CCR5-tropic and recoverable from HSPC populations that were depleted for stem cells . These unexpected results support recent studies showing that non-stem cell progenitors can persist in vivo for years and provide evidence that they may form a significant reservoir in HIV infected people . We also provide strong evidence that progenitor cells , including multipotent progenitors , harbor HIV receptors . These results are consistent with other studies investigating the lineage potential of CD4 subsets using functional assays [16 , 17 , 19] . Two studies showed that CD34+ CD4high and CD4low/- populations include clonogenic progenitors and Louache et al furthermore demonstrated that CD34+ CD4+ HSPCs are enriched for long-term culture-initiating cells [16 , 17] . Another study extended these results using human fetal liver to show that CD34+CD4+ cells are able to engraft in an immunodeficient mouse , unlike CD34+CD4- cells [19] . In addition , HIVs that require CD4 for entry are able to infect and express marker genes in HSCs based on a gold standard functional assay ( stable engraftment and generation of all hematopoietic lineages ) [3] . Thus , CD4 and other HIV receptors are expressed on hematopoietic progenitors . Preferential infection of the CD4high subset partially explains another study that was unable to detect provirus in HSPCs from infected people [20] . In this study , flow cytometry was used to isolate Lin−CD34+ CD4- cells , obtaining a mean purity of 76 . 7% that was substantially lower than the samples described here ( mean purity 94 . 1% for Sort 1 and 90 . 3% for Sort 2 ) . Based on the data presented here , removal of the CD4+ population would have removed the HSPC population most likely to be infected . In addition , this small study of 8 donors ( 3 initiating therapy during chronic infection and 5 initiating therapy during acute infection ) was underpowered to detect provirus in HSPCs . The authors estimate that in these 8 patients , if proviral genomes were present , their frequency would be 0 . 0003%–0 . 003% ( upper 95% confidence bounds ) . Given that 59% of our donors were positive and that the mean frequency of provirus in our cohort was 2 . 4 copies per million cells ( 0 . 0002%; range 18 to < 0 . 8 copies per million cells ) , the small study size and the small number of cells screened provide additional explanations for why this and another similarly powered study [21] were negative . Importantly , we isolated four near full-length genomes from HSPCs and a detailed analysis of open reading frames and cis-acting elements revealed they are likely to be functional . However , demonstration of functional virus using viral outgrowth assays will require additional studies using larger cell numbers . Studies in T cells have shown that only about a tenth of functional virus can be detected in outgrowth assays [22] . A Poisson analysis using a mean frequency of 3 copies of provirus per million HSPCs with 30% functional based on sequencing suggests 60 million HSPCs will be needed for 95% certainty of detecting one infectious unit . Given that we obtain about 2 . 5 million HSPCs for each donor from 20 cc of marrow , we would need to dramatically increase our aspiration size to acquire sufficient cells , which would not be easy to accomplish because of patient discomfort . The low rate of infection in HSPCs likely explains why an earlier study utilizing low numbers of HSPCs ( approximately one million ) yielded negative results in outgrowth assays [21] . In addition , while we have shown that transcriptionally latent viral genomes in HSPCs can be reactivated by TNFα and histone deacetylase inhibitors in vitro after cell culture [3 , 13] , studies using large cell numbers are needed to determine the optimal strategies to effectively reactivate proviral genomes to promote viral release from fully quiescent HSPCs tested ex vivo . Nevertheless , the conclusion that HIV indeed infects HSPCs was confirmed by the detection of clonal HIV proviral genomes in differentiated lineages that matched provirus from HSPCs . Because the differentiated cells were CD4-negative lineages and because the provirus contained signature inactivating deletions , these results can’t be explained by coincident infection . Moreover , we confirmed the presence of these genomes by isolating cell-associated mRNA containing the same deletion from activated CD4-negative cells . Further , we showed by phylogenetic analysis that the genomes frequently isolated from CD4-negative lineages formed a unique clonal population within the donor , indicating that contamination from other cell types was an unlikely explanation of our findings . In sum , the most likely explanation is that these genomes were transmitted to CD4-negative progeny through differentiation of a CD4-positive progenitor . In addition , we also detected a proviral genome with a unique signature deletion in both HSPC and CD4+ cells indicating that infected HSPCs can also differentiate into CD4+ cells . In most cases , detection of proviral genomes in CD4-negative lineages was rare with only a small number of proviral genomes detected per million cells screened . The exception was donor 431000 in which we detected a defective provirus at a higher frequency ( approximately one per 100 , 000 cells screened ) . Because replication competent virus could disrupt differentiation due to cytopathic effects , it is not surprising that viral spread from differentiating HSPCs would be uncommon with functional virus , occurring at a higher frequency in cells harboring a defective viral genome that might allow normal differentiation to occur . In addition , we detected proviral genomes more often in CD4-negative lineages from donors with CXCR4-tropic virus , consistent with its ability to target a wider range of HSPC subtypes , including MPPs and HSCs . With the exception of one donor ( 431000 ) , we did not find CCR5-tropic provirus in differentiated CD8 , B and NK lineages found in the peripheral blood , which is consistent with observations that CCR5-tropic HIV more commonly infects restricted myeloid progenitor cells [4] . Although HSCs are the main drivers for reconstitution of all hematopoietic lineages in xenograft models , new insights in animal and human disease models have shown contributions of non-stem cell progenitors to steady state hematopoiesis over long periods of time [6–8] . Non-stem cell progenitors appear to survive longer than previously thought in the bone marrow without contribution from HSCs , with non-stem cell clones sequentially recruited over time to produce mature blood cells [6–8 , 23 , 24] . Our data that CCR5-tropic provirus persists for years in non-stem cell progenitors is to our knowledge the first evidence that non-stem cell progenitors persist for years in humans without evidence of bone marrow disease . Given that non-stem cell progenitors persist , the prevalence of CCR5-tropic HIV in this compartment is not surprising . During acute infection when circulating virus peaks , the majority of virus is CCR5-tropic [25] . However , we also detected persistent provirus that encodes Env proteins capable of utilizing CXCR4 to enter cells . Assuming transmitting virus is nearly uniformly CCR5-tropic , as some studies have indicated , the presence of persistent reservoirs of CXCR4-tropic provirus may indicate that reservoirs continue to form during evolution to CXCR4 tropism in some donors . Overall , these results support a new model in which non-stem cell progenitors are important long term contributors to normal hematopoiesis and moreover that these cells can serve as a persistent reservoir for HIV provirus .
HIV-infected individuals were recruited through the University of Michigan HIV-AIDS Treatment Program and the Henry Ford Health System . Written informed consent was obtained according to a protocol approved by the University of Michigan Institutional Review Board and Henry Ford Institutional Review Board ( U-M IRB number HUM00004959 and HFH IRB number 7403 ) . Donors were >18 years old , with normal white blood cell counts and plasma viral loads were <48 copies/ml for at least 6 months on antiretroviral therapy . 100 ml of peripheral blood and 20 ml of bone marrow were obtained from each donor . All collected samples were coded . Whole umbilical cord blood ( CB ) from uninfected donors was obtained from the New York Blood Center and whole bone marrow was obtained commercially ( AllCells Ltd . ) . All collected samples were anonymized . For isolation of HSPCs , mononuclear cells were purified by Ficoll-Hypaque centrifugation and adherence depleted in serum-free StemSpan medium ( StemCell Technologies ) for 1–2 hours at 37°C . Sort 1 cells were isolated with a CD133 MicroBead Kit ( Miltenyi Biotec ) according to the manufacturer’s protocol , using two sequential sorts for increased purity . ( For donations 453000 , and 454304 , we used 1 . 5 times the recommended MicroBeads to increase yield . ) Sort 2 cells were isolated from the Sort 1 flow-through using EasySep Human CD34 Positive Selection Kit ( StemCell Technologies ) according to the manufacturer’s protocol , using two sequential sorts . Where indicated , lineage-positive cells were depleted using the EasySep Lineage Depletion Kit ( StemCell Technologies ) before proceeding to the CD133 magnetic sort . CD4 negative PBMCs from the human donors described in the ethics statement were purified by depletion of CD4+ cells with MicroBeads ( Miltenyi Biotec ) according to the manufacturer’s protocol modified for a bead:cell ratio of 1 . 5:1 and passage over two sequential LS magnetic columns . Depleted cells were stained and sorted as indicated in the text to remove residual CD4+ cells on a MoFlo Astrios flow cytometer . Supernatant and cell associated RNA was extracted using TRIzol LS and TRizol reagents , respectively according to the manufacturer’s protocols ( Invitrogen ) and converted to cDNA using qScript cDNA Supermix or qScript Flex cDNA Kit according to manufacturer’s instructions ( Quanta Biosciences ) . RNA from viral supernatants was quantified by real time PCR using TaqMan Fast Mastermix ( Applied Biosystems ) on an Applied Biosystems 7300 thermocycler using primers and probes as previously described [26] and used in SGA PCR described below . For gene expression analysis , bone marrow cells were isolated and harvested as described [27] . RNA was extracted from 3x104 double-sorted cells from each cell population . RNAseq was performed on the total RNA extracted from each cell population , adding equal amounts of 92 spiked-in RNA standards to each cell population . Since the amount of spiked-in RNA standards added to each sample was known , the relationship between RPKM ( reads per kilobase per million ) values and the number of transcripts for each spiked-in RNA could be determined by regression analysis [28] . RNAseq reads were aligned using Bowtie software [29] to NCBI build 37 ( mm9 ) of the mouse genome with the settings: -e 70 -k 1 -m 2—n 2 . The RPKM for each RefSeq gene and synthetic spike-in RNA was calculated using RPKM_count . py ( v2 . 3 . 5 ) counting only exonic reads ( -e option ) . Loess regression from R affy package was used to renormalize the RPKM values by using only the spike-in RNA to fit the loess with default parameters . Only the spike-in RNAs whose abundance could be robustly quantified ( RPKM values ≥ 1 ) were used in the loess normalization . DNA was prepared using the MagNA Pure Compact System ( Roche ) and used at limiting dilution for a 2-step SGA PCR validated for single copy sensitivity on ACH-2 cell DNA using primer sets shown in S5 Table . PBMC DNA from each donor was used to select the optimal primer sets for each donor . PCR assays were performed using a BioRad C1000 thermocycler as described in S6 Table . Amplicons were sequenced directly from the purified gel band . Virus was prepared by transfection of HIV or lentiviral genome containing plasmids into 293T ( ATCC ) cells as described [13] . Where indicated , the helper plasmid pCMV-HIV-1[30] and a plasmid encoding either VSV-G protein or an HIV envelope protein were used as described previously [3 , 31] . Intracellular Gag staining was performed as previously described [13] . Env expression vectors were generated using gel purified DNA cloned into pcDNA3 . 1/V5-His-TOPO TA . For phenotypic analysis of donor Env matching the HSPC V3 region , we used 3T3-CD4-CCR5 and CXCR4 cells ( NIH AIDS Reagent Repository ) [11] transduced with HIV-7/SF-GFP pseudotyped with the each Env and harvested for flow cytometry 3 days post-transduction . Antibodies to the following human proteins were used for flow cytometry: CD133 ( phycoerythrin [PE] conjugated; Miltenyi Biotec ) , CD34 ( conjugated with fluorescein isothiocyanate [FITC] , allophycocyanin [APC] , PE-Cy7; Miltenyi Biotec and eBioscience ) , CD3 ( APC; eBioscience or APC-H7-conjugated; BD Bioscience ) , CD4 ( clone OKT4 , unconjugated , Brilliant Violet 605 conjugated; BD Biosciences or AlexaFluor488 conjugated; eBioscience ) , CD45RA ( APC conjugated; eBioscience ) , CD38 ( PE-Cy7 conjugated; eBioscience ) , CD10 ( Biotin conjugated; eBioscience ) , HIV-1 Gag ( clone KC57 , FITC conjugated; Beckman Coulter ) , CD8 ( PE conjugated; BioLegend ) , CD19 ( APC conjugated; BD Bioscience ) , CD56 ( PE-Cy7 conjugated; eBioscience ) . and Human Hematopoietic Lineage Cocktail ( FITC conjugated; eBioscience ) . The secondary reagents used were streptavidin ( Brilliant Violet 421 conjugated; BD Biosciences ) and anti-mouse IgG2b ( Alexa Fluor 647 conjugated; Invitrogen ) . Non-viable cells were identified and excluded from sorts and analyses by staining with 7-aminoactinomycin D ( 7-AAD ) or DAPI . Samples were analyzed using a BD FacsCanto cytometer . Cell sorting was performed using a MoFlo XDP ( Beckman Coulter ) , MoFlo Astrios ( Beckman Coulter ) , FACSJazz ( BD Biosciences ) or FACSAria ( BD Biosciences ) flow cytometer . For flow cytometric analysis and isolation of specific murine hematopoietic progenitors for RNA isolation , cells were incubated with combinations of antibodies to the following cell-surface markers conjugated to FITC , PE , PerCP-Cy5 . 5 , APC , PE-Cy7 , or biotin: CD3ε ( 17A2 ) , CD4 ( GK1 . 5 ) , CD5 ( 53–7 . 3 ) , CD8α ( 53–6 . 7 ) , CD11b ( M1/70 ) , CD16/32 ( FcΥRII/III; 93 ) , CD34 ( RAM34 ) , CD43 ( 1B11 ) , CD45R ( B220; RA3-6B2 ) , CD48 ( HM48-1 ) , CD117 ( c-kit; 2B8 ) , CD127 ( IL7Rα; A7R34 ) , CD150 ( TC15-12F12 . 2 ) , Ter119 ( TER-119 ) , Sca1 ( D7 , E13-161 . 7 ) , Gr-1 ( RB6-8C5 ) , and IgM ( II/41 ) . For isolation of CD150+CD48-Lineage-Sca-1+c-kit+ ( CD150+CD48-LSK ) HSCs and CD150-CD48-LSK MPPs , lineage markers included CD3 , CD5 , CD8 , B220 , Gr-1 , and Ter119 . For isolation of CD34+CD16/32lowCD127-Sca-1-LK CMPs and CD34+CD16/32highCD127-Sca-1-LK GMPs , these lineage markers were supplemented with antibodies against CD4 and CD11b . Other sorted populations included Gr-1+ cells , IgM-CD43+B220+ pro-B cells , IgM-CD43-B220+ pre-B cells , and unfractionated bone marrow cells . | People who are effectively treated with antiretroviral medication harbor persistent forms of HIV that are integrated into the cellular genome . While HIV is cytopathic to most cells , transcriptionally silent , latent forms do not express toxic HIV gene products and can survive in the host for years . When conditions change , the latent virus can be activated to reinitiate infection . Because of the capacity for virus to spread , cure of HIV will require that we identify and eradicate all cells harboring functional HIV provirus . CD4+ T cells are abundant and easily identified as harboring proviral genomes . However , rare cell types that express HIV receptors , such as bone marrow hematopoietic progenitor and stem cells can also be infected by the virus potentially serving as barriers to cure strategies . We found that HIV can infect and persist in progenitor sub-types that were previously thought to be short lived , which expands the types of cells that can support reservoir formation . In addition , we found that HIV can spread by proliferation and cellular differentiation without the need for viral gene expression and virion production that could reveal the infection to the immune system . A deeper understanding of viral reservoirs is critically important for developing strategies that will succeed in viral eradication . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"blood",
"cells",
"flow",
"cytometry",
"hiv",
"infections",
"medicine",
"and",
"health",
"sciences",
"immune",
"cells",
"pathology",
"and",
"laboratory",
"medicine",
"pathogens",
"immunology",
"microbiology",
"cell",
"differentiation",
"retroviruses",
"viruses",
"immunodeficiency",
"viruses",
"bone",
"marrow",
"cells",
"developmental",
"biology",
"rna",
"viruses",
"stem",
"cells",
"molecular",
"biology",
"techniques",
"research",
"and",
"analysis",
"methods",
"infectious",
"diseases",
"spectrum",
"analysis",
"techniques",
"white",
"blood",
"cells",
"artificial",
"gene",
"amplification",
"and",
"extension",
"animal",
"cells",
"medical",
"microbiology",
"hiv",
"microbial",
"pathogens",
"t",
"cells",
"molecular",
"biology",
"spectrophotometry",
"cytophotometry",
"cell",
"biology",
"polymerase",
"chain",
"reaction",
"viral",
"pathogens",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"viral",
"diseases",
"lentivirus",
"organisms"
] | 2017 | CD4 is expressed on a heterogeneous subset of hematopoietic progenitors, which persistently harbor CXCR4 and CCR5-tropic HIV proviral genomes in vivo |
The continued evolution of bacterial pathogens has major implications for both human and animal disease , but the exchange of genetic material between host-restricted pathogens is rarely considered . Streptococcus equi subspecies equi ( S . equi ) is a host-restricted pathogen of horses that has evolved from the zoonotic pathogen Streptococcus equi subspecies zooepidemicus ( S . zooepidemicus ) . These pathogens share approximately 80% genome sequence identity with the important human pathogen Streptococcus pyogenes . We sequenced and compared the genomes of S . equi 4047 and S . zooepidemicus H70 and screened S . equi and S . zooepidemicus strains from around the world to uncover evidence of the genetic events that have shaped the evolution of the S . equi genome and led to its emergence as a host-restricted pathogen . Our analysis provides evidence of functional loss due to mutation and deletion , coupled with pathogenic specialization through the acquisition of bacteriophage encoding a phospholipase A2 toxin , and four superantigens , and an integrative conjugative element carrying a novel iron acquisition system with similarity to the high pathogenicity island of Yersinia pestis . We also highlight that S . equi , S . zooepidemicus , and S . pyogenes share a common phage pool that enhances cross-species pathogen evolution . We conclude that the complex interplay of functional loss , pathogenic specialization , and genetic exchange between S . equi , S . zooepidemicus , and S . pyogenes continues to influence the evolution of these important streptococci .
Streptococcus equi subspecies equi ( S . equi ) is the causative agent of equine strangles , characterized by abscessation of the lymph nodes of the head and neck . Rupture of abscesses formed in retropharyngeal lymph nodes into the guttural pouches leads to a proportion of horses becoming persistently infected carriers . These carriers transmit the organism to naïve horses and play an important role in disease spread . S . equi is believed to have evolved from an ancestral strain of Streptococcus equi subspecies zooepidemicus ( S . zooepidemicus ) [1] , [2] , which is associated with a wide variety of diseases in horses and other animals including humans . Both of these organisms belong to the same group of streptococci as the human pathogen Streptococcus pyogenes . Previous work has shown that S . equi produces four superantigens ( SeeH , SeeI , SeeL and SeeM ) [3]–[5] , two secreted fibronectin-binding proteins ( SFS and FNE ) [6] , [7] , a novel M-protein ( SeM ) [8] , an H-factor-binding protein ( Se18 . 9 ) [9] and a novel non-ribosomal peptide synthesis system [10] , but little is known about other factors that influence differences in the virulence of these closely related streptococci . We determined the complete genome sequence of S . equi strain 4047 ( Se4047 ) , a virulent strain isolated from a horse with strangles in the New Forest , England , in 1990 [11] and S . zooepidemicus strain H70 ( SzH70 ) , isolated from a nasal swab taken from a healthy Thoroughbred racehorse in Newmarket , England , in 2000 [2] . Using comparative genomic analysis to identify Se4047-specific loci , and subsequent screening of S . equi and S . zooepidemicus strains from around the world , we provide evidence of the genetic events that have shaped the evolution of the S . equi genome , and led to its emergence as a host-restricted pathogen .
Multilocus sequence typing ( MLST ) has provided evidence of the close genetic relationship of S . equi and S . zooepidemicus [2] . The genomes of Se4047 ( ST-179 ) and SzH70 ( ST-1 ) support the overall relatedness , but also reveal evidence of genome plasticity that has generated notable diversity . The two genomes are similar in size: the Se4047 genome consists of a circular chromosome of 2 , 253 , 793 bp ( Figure 1A ) encoding 2 , 137 predicted coding sequences ( CDSs ) , and the SzH70 genome contains a chromosome of 2 , 149 , 866 bp ( Figure 1B ) , encoding 1 , 960 predicted CDSs . Much of the Se4047 genome is orthologous to the SzH70 genome: 1671 Se4047 CDSs have SzH70 orthologs . Of the remaining 466 non-orthologous Se4047 CDSs , 422 are found on mobile genetic elements ( MGEs; for details of the regions of variation in the Se4047 and SzH70 genomes see Table S1 ) . Recently , the genome sequence of S . zooepidemicus strain MGCS10565 ( SzMGCS10565 ) was published [12] . This strain was isolated from a human case of nephritis that was part of a severe epidemic in Brazil [13] . MLST ( http://pubmlst . org/szooepidemicus/ ) analysis indicates that SzH70 and SzMGCS10565 ( ST-72 ) are genetically distinct from each other and Se4047 . Comparative analysis reveals that the number of orthologs in the Se4047 genome is slightly higher for SzH70 ( 78 . 2% ) than for SzMGCS10565 ( 77 . 4% ) ; 76 . 3% of the Se4047 CDSs have matches in both S . zooepidemicus strains . For the purposes of this study we have primarily focused our analysis on the comparison of equine isolates , Se4047 with SzH70 . The chromosomes of Se4047 and SzH70 are generally collinear except for two inversions around the origin of replication ( Figure 2 ) . The smaller central inversion is due to recombination events in Se4047 between identical ISSeq3 elements on opposite replichores . The larger rearrangement is due to an inter-replichore inversion in SzH70 of unknown origin ( Figure 2 ) . Both the Se4047 and SzH70 genomes contain two copies of hasC which encode UDP-glucose pyrophosphorylases [14] . In SzH70 one copy of hasC ( SZO17510 ) has been translocated to the opposite replichore by the previously mentioned large reciprocal inversion . There is also a small intra-replichore inversion ( ∼14 kb ) in Se4047 between the two copies of hasC ( SEQ0271 and SEQ0289 ) . The hasC-mediated inversion in Se4047 rearranges the genes associated with capsule production [14] and may explain why S . equi produces such high levels of hyaluronate capsule . Comparison of the predicted functions of the genes encoded in the Se4047 and SzH70 genomes revealed that Se4047 has the same number , or fewer CDSs , in each of the functional classes with the exception of protective responses and adaptation and laterally acquired elements ( Figure 3A ) . The number of pseudogenes in Se4047 is also elevated in comparison to SzH70 . The additional protective response and adaptation CDSs in Se4047 are associated with the biosynthesis of a putative siderophore [10] , and are carried on a MGE region of the genome ( ICESe2; Figure 1 ) . The relative expansion of laterally acquired elements , and increased number of pseudogenes in Se4047 suggests that the evolution of S . equi has been shaped by recent gene loss and gain . A corollary of this genome plasticity appears to have been a reduction in ancestral capabilities , and the introduction of novel functions , which have enabled S . equi to exploit a new niche . Se4047 has 58 partially deleted genes and 78 pseudogenes , compared with 62 and 29 respectively in SzH70 ( Figure 3B and Table S1 ) . In particular , Se4047 is enriched for mutations associated with catabolic metabolism , transport , and the cell envelope . Such gene loss is typical of other host-restricted bacteria that have evolved from versatile ancestors [15] , [16] . The loss of ancestral functions appears to have played a seminal role in the evolution of S . equi , resulting in a refinement of its nutritional capabilities , and its host-cell interactions . Carbohydrate metabolism in streptococci plays an important role in colonization of mucosal surfaces [17] . Carbohydrate fermentation is also commonly used to differentiate S . equi strains from S . zooepidemicus [18] . Comparison of the genome sequences identified a 5 kb deletion in the Se4047 genome that partially deleted lacD and lacG and deleted lacE , lacF and lacT . Se4047 also contains a deletion of sorD immediately upstream of SEQ0286 and a deletion between SEQ0536 and SEQ0537 that spans the operon required for ribose fermentation . Specialization of S . equi has probably rendered these pathways redundant , resulting in their loss . To determine if differences in gene content identified through genome comparison represented variation between S . equi and S . zooepidemicus or variation within their populations , we screened by PCR a panel of S . equi and S . zooepidemicus strains that are representative of the wider population as defined by MLST [2] . This included 26 isolates of S . equi ( representing 2 STs ) and 140 isolates of S . zooepidemicus ( representing 95 STs ) [2] . All 26 S . equi strains examined lacked lacE , sorD and rbsD and the capacity to ferment lactose , sorbitol or ribose . However , only 15 ( ST-7 , ST-39 , ST-57 , ST-97 and ST-106 ) and 1 ( ST-39 ) of 140 S . zooepidemicus isolates tested did not ferment ribose or sorbitol , respectively ( Figure 4 ) . Hyaluronate lyases are secreted enzymes that degrade hyaluronic acid and chondroitins facilitating invasion by bacteria and their toxins [19] . The SzH70 genome contains a single CDS encoding a putative hyaluronate lyase ( SZO06680 ) . However , the Se4047 orthologue , SEQ1479 , contains a 4 bp deletion ( TCTC ) leading to a frameshift at codon 199 . Se4047 has acquired a different hyaluronate lyase ( SEQ2045 ) encoded on a prophage . This type of phage-encoded enzyme typically has much lower activity and reduced substrate range [20] than orthologues of SZO06680 [21] and may provide an explanation for why S . equi infection rarely progresses beyond the lymphatic system . The 4 bp deleted in strain Se4047 was also absent in all 26 strains of S . equi tested , whereas these 4 bp were present in all strains of S . zooepidemicus . However , one strain of S . zooepidemicus ( ST-57 ) was PCR negative due to an IS element insertion 905 bp from the translational start ( Figure 4 ) . Reduced hyaluronate lyase activity provides an alternative explanation as to why S . equi maintains high levels of hyaluronate capsule and in agreement with this , the ST-57 isolate of S . zooepidemicus that tested PCR negative also maintained high levels of capsule . Increased levels of capsule may enhance resistance to phagocytosis [22] , but could also reduce adhesion to the mucosal surface [23] . To demonstrate that mucoid colony phenotype was due to hyper-encapsulation , we grew Se4047 , SzH70 and the ST-57 isolate on plates containing hyaluronidase [24] . All colonies were no longer mucoid in appearance and resembled SzH70 ( Figure S1 ) . Gram-positive bacteria classically display an array of cell wall-anchored proteins on their surface , which are attached covalently through a process mediated by sortase enzymes [25] . In many cases , these cell wall-anchored proteins have been shown to play a role in modulating host-cell interactions . Two putative sortase CDSs are present in the Se4047 genome: srtA ( SEQ1171 ) and srtC . 1 ( SEQ0937 ) , whereas the SzH70 genome contains five: srtA ( SZO09440 ) , srtC . 1 ( SZO11490 ) , srtC . 2 ( SZO18270 ) , srtC . 3 ( SZO18280 ) and srtC . 4 ( SZO18290 ) . Together , these enzymes potentially process 29 and 39 putative surface proteins in Se4047 and SzH70 , respectively ( Table S2 ) . S . equi strains typically bind significantly lower quantities of fibronectin than those of S . zooepidemicus [7] . One possible explanation for this is a one-base deletion within SEQ0375 ( fne ) that was conserved in all strains of S . equi examined [7] . The base deletion in fne results in the loss of an LPXTG surface anchor and leads to the production of a secreted product , FNE , which binds both fibronectin and collagen [7] . Reduction in the fibronectin-binding properties of Staphylococcus aureus increases virulence in a rat pneumonia model [26] and truncation of fne has been proposed to increase the virulence of S . equi [7] . Our analysis identifies other examples of mutation and gene loss that are likely to contribute to decreased fibronectin binding in S . equi . The surface protein Shr of S . pyogenes binds heme and transfers it to the streptococcal heme-binding protein Shp for import by the HtsABC heme transporter [27] . Shr also binds fibronectin and contributes to attachment of S . pyogenes to epithelial cells [28] . SEQ0443 encoding Shr in S . equi contains a frameshift mutation after codon 442 that truncates this protein . Pili play an important role in the adherence of S . pyogenes to host tissues [26] . The SzH70 genome contains two loci that encode genes required putatively for pilus expression . The first of these ( SZO11490-SZO11520 ) shares 84–96% amino acid sequence identity to SEQ0934-SEQ0937 of Se4047 and 94–99% amino acid identity with the FimI locus of the recently published human disease isolate SzMGCS10565 [12] . However , the tetR-like regulator SEQ0934 of Se4047 contains a nonsense mutation at codon 43 that may lead to constitutive pilus production , longer pili that could more effectively protrude through the larger capsule of S . equi [29]–[31] and increased collagen-binding [32] . The second SzH70 pilus locus consists of CDSs encoding three putative sortase enzymes , SrtC . 2 , SrtC . 3 and SrtC . 4 , one putative exported protein ( SZ18300 ) and three putative surface proteins ( SZO18310-SZO18330 ) , which share 58% , 76% and 68% amino acid sequence identity with Spy0117 , Spy0116 and the fibronectin-binding protein Spy0115 of S . pyogenes MGAS10750 , respectively [33] and an AraC-like transcriptional regulator ( SZO18340 ) . The genome of strain Se4047 lacks this putative pilus locus through an ISSeq3 element-mediated deletion . None of the 26 isolates of S . equi , but 81 of 140 S . zooepidemicus isolates tested positive for srtC . 2 or srtC . 3 by PCR . The genome of SzMGCS10565 does not contain a homologue of this SzH70 pilus locus , but instead contains two other consecutive pilus loci Fim II and Fim III at the same genome location . Fim III is flanked by an AraC-like regulator ( Sez_1830 ) , which is orthologous to SZO18340 of SzH70 . Diversification of pilus loci could play an important role in the ability of S . zooepidemicus strains to infect different hosts and tissues . The SzH70 and SzMGCS10565 genomes encode a 131 kDa putative surface protein containing 1 , 160 amino acids with an LPXTG motif ( SZO08560 and Sez_1114 ) . However , the Se4047 genome encodes only the final 112 amino acids of this protein ( SEQ1307a ) and lacks an adjacent gene predicted to encode a recombinase ( SZO08550 and Sez_1116 ) . SZO08560 and Sez_1114 share sequence similarity with hypothetical proteins of S . suis strain 05ZYH33 ( SSU05_0473 ) and S . agalactiae strain COH1 ( SAN_1519 ) and contain four Listeria-Bacteroides repeat Pfam domains ( PF09479 ) . The ∼70 amino acid residue repeats occur in a range of Gram-positive surface proteins including the InlA internalin of Listeria monocytogenes [34] ( Figure S2 ) . InlA interacts with E-cadherin to promote invasion of L . monocytogenes into particular host cells [35] . Examination of the SzH70 genome sequencing data revealed five sequence reads that positioned the promoter region of SZO08560 ( −170 bp to −55 bp ) in the reverse orientation . This sequence is bordered by GTAGACTTTA and TAAAGTCTAC inverted repeats and we propose that inversion of this sequence switches transcription of SZO08560 on or off , thereby modulating the production of this surface protein in a manner akin to phase variation in E . coli ( Figure 5 ) [36] . Reverse transcription qPCR using RNA extracted from log-phase cultures of SzH70 and normalized for expression of the housekeeping gene gyrA demonstrated that the SZO08560 promoter of SzH70 transcribed 44-fold more RNA in the forward direction than the reverse . To our knowledge this is the first potential example of recombinase regulation of surface protein production in streptococci . None of the 26 isolates of S . equi , but 101 of 140 S . zooepidemicus isolates tested positive for SZO08560 by PCR . SzMGCS10565 contains an IS element between the inverted repeats bordering the Sez_1114 promoter and the recombinase ( Sez_1116 ) , the consequences of this on transcription of Sez_1114 are not yet known . Ess ( ESAT-6 secretion system ) specialized secretion systems have been identified in Mycobacterium tuberculosis and S . aureus and shown to trigger cell-mediated immune responses including IFN-gamma production that play an important role in virulence [37] . The SzH70 genome contains a cluster of 9 genes ( SZO14600-SZO14680 ) with similarity to the Ess of SzMGCS10565 and S . aureus [37] . Unexpectedly , the Se4047 genome lacks esaA , esxA and part of SEQ0576 associated with the presence of an upstream ISSeq3 element ( SEQ0575-SEQ0574 ) . PCR analysis showed that none of the 26 strains of S . equi , but 138 of 140 S . zooepidemicus strains examined contained the esaA gene ( Figure 4 ) . The increased size of the Se4047 genome compared to the genome of SzH70 is due to the acquisition of a large number of MGEs . Together these make up a total of 16 . 4% of the Se4047 genome . In contrast 7 . 5% of the SzH70 genome is composed of MGEs . Several of the MGEs in Se4047 carry notable virulence determinants absent in SzH70 . The acquisition of these regions by a progenitor may have opened up new pathogenic niches , and been critical in the emergence of S . equi . Unlike SzH70 , Se4047 is polylysogenic , containing 4 prophage . The acquisition of prophage plays an important role in the evolution of many pathogenic bacteria [38] . Cargo genes carried by prophage can increase the survival fitness or enhance niche adaptation of the lysogen [38] , [39] . Phage repressor and superinfection exclusion functions also confer a selective advantage to the lysogen by providing immunity against lytic infection [40] . Comparison of the sequences of each of the prophage found in Se4047 with each other showed only limited mosaic similarity . However , comparison with prophage sequences in the public databases revealed more extensive similarity with prophage from S . pyogenes , so much so that clustering analysis has demonstrated that the individual S . equi prophage are more related to phage in the other sequenced S . pyogenes genomes than they are to each other ( Figure 6 ) , suggesting commonality in the phage pool of these pathogens . The first of the four Se4047 prophage , φSeq1 , is 39 kb in size , contains the CDSs SEQ0133-SEQ0197 and is integrated immediately after the tRNA-Cys gene . The CDSs of φSeq1 do not have homology to known virulence factors . However , re-circularized φSeq1 was amplified by PCR and confirmed by sequencing across the join of the re-circularized phage following preparation of phage particles present in cultures of Se4047 treated with mitomycin C . Release of φSeq1 may result in killing of susceptible bacteria , such as S . zooepidemicus , which may compete to colonize the epithelium of the equine nasopharynx , thereby resulting in more efficient attachment of S . equi and its invasion of the lymphatic system . Such a mechanism is also seen in a lysogen of Salmonella enterica serovar Typhimurium , which releases low titers of phage that lysed competing non-lysogenic strains [40] . The 41 kb φSeq2 ( SEQ0787-SEQ0851 ) is integrated into the putative C-terminal sequence of an ATP-dependent DNA helicase ( SEQ0786 ) and contains a CDS ( SEQ0849 ) that shares 98% predicted amino acid sequence identity with the phospholipase A2 , SlaA , of S . pyogenes M3 MGAS315 [38] . SlaA is known to contribute to streptococcal virulence and its recent acquisition by S . pyogenes M3 ( in approximately 1987 ) was associated with increased morbidity and mortality [38] . Deletion of slaA reduced the virulence of S . pyogenes in a mouse intraperitoneal infection model and severely compromised its ability to colonize the upper respiratory tract of a macaque model of pharyngitis [41] . slaA is widely distributed amongst S . zooepidemicus STs ( 44 of 140 isolates ) and all 26 strains of S . equi tested here ( Figure 4 ) . Of particular note was the observation that S . equi CF32 , which was isolated from a horse with strangles during 1981 , contained slaA . This isolate predates all slaA positive isolates of S . pyogenes [42] , and it is possible that slaA in the S . pyogenes gene pool gene may have arisen via phage-mediated horizontal transfer from a slaA-containing strain of S . zooepidemicus or S . equi , although the precise evolutionary origins remain unclear . We were unable to detect re-circularized φSeq2 in phage particle preparations of Se4047 following mitomycin C treatment . However , this prophage appears to be intact and could re-circularize in response to other stimuli such as UV irradiation or heat shock . In support of the important role that these toxins may play in colonization and virulence of S . equi and S . zooepidemicus , we identified a gene encoding a second putative phospholipase A2 toxin , SlaB , sharing 70% amino acid sequence identity with SlaA of S . pyogenes in the genomes of Se4047 ( SEQ2155 ) and SzH70 ( SZO18670 ) . This gene , also identified in SzMGCS10565 ( Sez_1876 ) , was associated with the remnants of a hypothetical prophage gene and was present in all strains of S . equi and S . zooepidemicus tested ( Figure 4 ) . The 30 kb φSeq3 is integrated into SEQ1725 which encodes a putative late competence protein and contains CDSs SEQ1727-SEQ1765 including two cargo CDSs encoding the superantigens SeeL and SeeM , which share 97% and 96% amino acid sequence identity with SpeL and SpeM of S . pyogenes MGAS8232 , respectively [3] , [5] . The genes encoding SeeL and SeeM were present in all strains of S . equi and 4 of 140 isolates of S . zooepidemicus tested ( Figure 4 ) . Interestingly , these S . zooepidemicus isolates represented 3 unrelated STs ( ST-106 , ST-118 and ST-120 ) recovered from the same outbreak of equine respiratory disease in 1996 . S . equi CF32 also contained these superantigen genes , and predates SpeL- and SpeM-producing strains of S . pyogenes [43] , providing further evidence that S . equi and S . zooepidemicus act as reservoirs of virulence genes that may be transferred by lateral gene transfer events . Re-circularized φSeq3 was not detected by PCR of mitomycin C induced phage particle preparations of Se4047 . However , the CDSs of this prophage appear to be intact and may permit re-circularization in response to other stimuli . Finally , the 40 kb φSeq4 is inserted next to SEQ2035 , resulting in the truncation of this putative transcriptional repressor . φSeq4 contains cargo CDSs encoding the previously described superantigens SeeH ( SEQ2036 ) and SeeI ( SEQ2037 ) , which share 98% and 99% amino acid sequence identity with SpeH and SpeI , respectively [4] . Interestingly , φSeq4 was very closely related to φMan3 of S . pyogenes Manfredo ( Figure 7 ) . Although seeH and seeI were present in all strains of S . equi tested , we have not yet identified any strains of S . zooepidemicus that contain these genes . However , re-circularized φSeq4 was detected by PCR and confirmed by sequencing across the join of the re-circularized phage purified from cultures of Se4047 treated with mitomycin C . Our data suggest that the acquisition of φSeq4 by S . equi , possibly originating from a strain of S . pyogenes , may have been a very recent event that could have influenced the emergence of S . equi . To provide functional evidence for the production of superantigens by different strains of S . equi and S . zooepidemicus , we assayed the culture supernatants of our strain collection . This confirmed that all strains of S . equi and the strains of S . zooepidemicus containing seeL and seeM possessed significant mitogenic activity . However , the supernatants of 25 additional strains of S . zooepidemicus also had mitogenic activity . Several of these strains were related genetically by MLST , and clustered into three groups ( ST-123 , ST-127 and ST-141; ST-7 , ST-48 , ST-70 , ST-5 and ST-53; ST-8 , ST-46 and ST-113 ) ( Figure 4 and Table S3 ) . We propose that these strains probably contain genes encoding other S . pyogenes superantigens or novel genes that represent an additional reservoir of as yet uncharacterized superantigens . The absence of prophage in the SzH70 genome , and low frequency of phage associated superantigens in the screening of S . zooepidemicus strains , is in stark contrast to S . equi . One explanation for the lack of prophage in S . zooepidemicus is that systems exist in naturally transformable streptococci that provide resistance to uptake and incorporation of foreign DNA and may co-incidentally prevent stable prophage integration [12] . Se4047 lacks 9 putative competence genes ( Table S1 ) that are intact in SzH70 and SzMGCS10565 , which could provide an explanation for the polylysogenic nature of Se4047 . An alternative explanation of the proliferation of prophage in S . equi can be found in the genome comparison between SzH70 and Se4047 . In the SzH70 genome a locus containing a clustered regularly interspaced short palindromic repeat ( CRISPR ) array and CRISPR-associated ( CAS ) genes ( SZO14370-SZO14430 ) was identified , which has been deleted from the Se4047 genome due to recombination between ISSeq11 elements ( Table S1 ) . CRISPR arrays are composed of direct repeats that are separated by similarly-sized non-repetitive spacers . These arrays , together with a group of associated proteins , confer resistance to phage directed by sequence similarity between the spacer regions and the phage in question , possibly via an RNA-interference-like mechanism [44] , [45] . The SzH70 CRISPR contains eighteen spacer sequences , of which ten have no significant database matches , three share >94% identity with prophage sequences present in the published genomes of S . pyogenes , four spacers have identical matches with prophage sequences found in the Se4047 genome ( #6 with SEQ0163 , #7 with SEQ1743 , #8 with SEQ1745 and #15 with SEQ1727 ( seeM ) ) and one spacer ( #18 ) has a near identical match with the Se4047 prophage CDS SEQ0190 , differing only at the first nucleotide ( C to T ) . This latter spacer is the only exact match with the spacer sequences of SzMGCS10565 CRISPRs ( spacer 9 of CRISPR I ) [12] . The limited spacer similarity of SzH70 and SzMGCS10565 may reflect exposure to different phage in their respective host environments . The CRISPR loci of SzH70 and SzMGCS10565 may assist the development of resistance to circulating phage and maintain genome integrity . The CRISPR region of SzH70 was present in 93% ( 131/140 ) of S . zooepidemicus isolates examined by PCR , but was absent from all strains of S . equi tested ( Figure 4 ) . Deletion of the CRISPR locus from the ancestor of Se4047 is likely to have resulted in increased genome instability and illustrates that in some circumstances gene loss may in turn influence the subsequent rate of gene gain . Both the SzH70 and Se4047 genomes contain distinct integrative conjugative element ( ICE ) regions . This type of MGE element has been shown to be widely distributed [46] , and associated with the transfer of a diverse range of functions . One of the ICE in the Se4047 genome , ICESe2 , contained CDSs ( SEQ1233-SEQ1246 ) with similarity to the non-ribosomal peptide synthesis ( NRPS ) system of Clostridium kluyveri and Yersinia sp . that produce an unnamed siderophore [47] and the ferric iron-binding siderophore yersiniabactin [48] , respectively . We have demonstrated that the S . equi NRPS operon is required for the production of an undefined secreted molecule , provisionally named equibactin , which enhances the ability of S . equi to acquire iron [10] . Siderophore biosynthesis has not previously been identified in any streptococci [49] . However , homologues of SEQ1246 and SEQ1243 ( present as a pseudogene ) are in the genome of S . agalactiae NEM316 serotype III , suggesting that a locus with similarity to the S . equi NRPS operon may have been important to this organism at some time . The ICESe2 locus was present in all of the S . equi isolates , but in none of the diverse collection of S . zooepidemicus isolates examined ( Figure 4 ) . Given the importance of iron acquisition to other streptococcal pathogens [50] , the acquisition of ICESe2 may have contributed significantly to the increased pathogenesis of this Streptococcus . In particular , we hypothesize that more efficient acquisition of iron could enhance the ability of S . equi to generate lymph node abscessation , which is critical to the establishment of long term carriage and vital to the success of this bacterium . It is intriguing to note that the production of yersiniabactin by Y . pestis is essential to its virulence [51] . It will be important to determine the contribution of ICESe2 to the formation of abscesses in the lymph nodes of horses . A facet of the Se4047 genome suggestive of recent niche adaptation is the large increase in the number of IS elements relative to SzH70 ( SzH70 contains 30 whereas Se4047 contains 73; Table S4 ) . In particular there appears to have been an expansion of the IS3-family IS element , ISSeq3: the Se4047 genome contains 40 copies of ISSeq3 whereas SzH70 contains 4 ( ISSzo3 ) . An expansion of IS elements has been observed in several host-restricted pathogens , which have recently evolved from generalist ancestors [15] , [16] . An evolutionary consequence of niche transit is hypothesized to be that many genes become dispensable , allowing increased inactivation . Niche change is also associated with significant evolutionary bottlenecks , which will be enhanced by repeated acquisition of mobile genetic elements . This leads to small effective population sizes , resulting in lower efficiency of selection , which in turn allows gene mutation and expansion of IS elements through accelerated genetic drift . A corollary of the IS proliferation has been the loss of genes by deletion [15]: several of the previously described examples of gene loss ( eg . pilus locus and CRISPR locus ) probably occurred through insertion and recombination between IS elements ( Table S1 ) . The comparison of the genomes of Se4047 and SzH70 provides strong evidence that S . equi has passed through a genetic bottleneck during its evolution from an ancestral S . zooepidemicus strain . We have identified several examples of gene loss that serve to reduce the ancestral capabilities of S . equi and increase the opportunity for genetic change . The acquisition of new mobile genetic elements has been critical to the evolution of S . equi . However , surveillance of the S . zooepidemicus population has identified examples of strains that did not cause strangles , but contain genes encoding phospholipase A2 toxins and superantigens . Therefore , we propose that the key speciation event in the evolution of S . equi was the acquisition of ICESe2 , containing a novel NRPS involved in the acquisition of iron , which is the first of its kind to be identified in streptococci . The proposed functional effects that result from the genetic events highlighted by our analysis are summarized in Figure 8 . Our study provides strong evidence for genetic exchange between S . equi , S . zooepidemicus and S . pyogenes , which continues to influence the pathogenicity of these important bacteria . The genetic diversity of the S . zooepidemicus population as measured by MLST [2] suggests that further investigation of this species will be likely to identify many more genes of importance to both veterinary and human disease .
Se4047 was isolated from a horse with strangles in the New Forest , England , in 1990 [11] , and has been typed as ST-179 by MLST [2] . SzH70 was isolated from a nasal swab taken from a healthy Thoroughbred racehorse in Newmarket , England , in 2000 , and has been typed as ST-1 by MLST [2] . Details of all of the isolates examined in this study are presented in Table S3 and are also available on the online MLST database ( Available: http://pubmlst . org/szooepidemicus/ . Accessed 3 October 2008 ) . For the preparation of DNA for whole genome sequencing Se4047 and SzH70 were grown overnight in Todd Hewitt broth ( THB ) at 37°C in a 5% CO2 enriched atmosphere . Cells were harvested and chromosomal DNA was extracted according to the method of Marmur [52] with the addition of 5000 units of mutanolysin ( Sigma ) and 20 µg of RNaseA ( Sigma ) during the lysis step . For the study of hyaluronate capsule degradation strains were grown overnight on COBA strep select plates ( bioMérieux ) at 37°C in a 5% CO2 enriched atmosphere , with and without pre-absorption of plates with 50 µl of 40 mg ml−1 hyaluronidase ( Sigma cat# H2126 ) . The genome of Se4047 was obtained with ∼8× coverage from m13mp18 and pUC18 genomic shotgun libraries ( with insert sizes of 1 . 4 to 4 kb ) using big-dye terminator chemistry on ABI3700 automated sequencers . Large insert BAC libraries ( pBACe3 . 6 , with insert sizes of 10–20 kb; and pEpiFos1 , with insert sizes of 38–42 kb ) were used as scaffolds . The SzH70 genome was obtained with ∼8× coverage from pUC18 and pMAQ1b genomic shotgun libraries ( with insert sizes of 2–6 kb ) using big-dye terminator chemistry on ABI3700 automated sequencers . A large insert pBACe3 . 6 library ( with insert sizes of 20–23 kb ) was used as a scaffold . Repeats were bridged by read-pairs or end-sequenced PCR products . The sequence was finished and annotated as described previously using Artemis software to collate data and facilitate annotation [53] . Comparison of the genome sequences was facilitated by using the Artemis Comparison Tool ( ACT ) [54] . Orthologous proteins were identified as reciprocal best matches using FASTA [55] with subsequent manual curation . Orthology inferred from positional information was investigated using ACT . Pseudogenes had one or more mutations that would prevent correct translation; each of the inactivating mutations was subsequently checked against the original sequencing data . The sequence and annotation of the Se4047 and SzH70 genomes have been deposited in the EMBL database under accession numbers FM204883 and FM204884 respectively . Sequences used for comparative genomic analysis were: S . zooepidemicus MGCS10565 ( CP001129 ) [12] , S . uberis 0140J ( AM946015 ) [56] , S . pyogenes Manfredo ( AM295007 ) [57] , S . thermophilus CNRZ1066 ( CP000024 ) [58] , S . suis P1/7 ( http://www . sanger . ac . uk/Projects/S_suis/ ) ( Holden et al . , unpublished ) , S . pneumoniae TIGR4 ( AE005672 ) [59] , S . sanguinis SK36 ( CP000387 ) [60] , S . mutans UA159 ( AE014133 ) [61] , S . agalactiae NEM316 ( AL732656 ) [62] , S . gordonii str . Challis substr . CH1 ( CP000725 ) [63] and Lactococcus lactis subsp . lactis IL1403 ( AE005176 ) [64] . S . pyogenes prophage sequences were extracted from the genomes of S . pyogenes strains Manfredo ( AM295007 ) [57]; SSI-1 ( BA000034 ) [65] , SF370 ( AE004092 ) [66] , MGAS315 ( AE014074 ) [42] , MGAS8232 ( AE009949 ) [59] , MGAS10394 ( CP000003 ) [60] , MGAS6180 ( CP000056 ) [61] , MGAS5005 ( CP000017 ) [67] , MGAS2096 ( CP000261 ) [33] , MGAS9429 ( CP000259 ) [33] , MGAS10270 ( CP000260 ) [33] and MGAS10750 ( CP000262 ) [33] . Prophage CDSs were clustered into homology groups using TribeMCL ( Centre for Mathematics and Computer Science and EMBL-EBI ) [68] with a cut-off of 1e−50 . The ability of isolates to ferment lactose , ribose and sorbitol was determined in Purple broth ( Becton Dickinson ) as previously described [18] . Equine PBMC were purified from heparinised blood by centrifugation on a Ficoll density gradient . PBMC were incubated with S . equi or S . zooepidemicus culture supernatants diluted 1/20 . PBMC proliferation was detected by overnight incorporation of 3H thymidine after 3 days of culture . Equine PBMC proliferation is expressed as stimulation index ( SI ) calculated as follows ( experimental response/control response ) . A SI≥2 was considered as positive . Genomic DNA from a diverse set of 26 S . equi strains and 140 S . zooepidemicus strains was prepared from single colonies grown on COBA strep select plates ( bioMérieux ) and purified using GenElute spin columns according to manufacturer's instructions ( Sigma ) . The relatedness of MLST STs was determined using ClonalFrame [69] . Gene prevalence was then determined by quantitative PCR ( QPCR ) using a SYBR green based method with a Techne Quantica instrument . For the QPCR , 10 ng DNA diluted was mixed with 0 . 3 µM forward and reverse primers ( Table S5 ) and 1× ABsolute QPCR SYBR green mix ( Abgene ) in a total volume of 20 µl and subjected to thermocycling at 95°C for 15 min , followed by 40 cycles of 95°C for 15 s , 55°C for 30 s and 72°C for 30 s . Dissociation curves were analyzed following a final ramp step from 60°C to 90°C with reads at 0 . 5°C increments to rule out non-specific amplification . Data were analyzed using Quansoft software ( Techne ) . Crossing point values relative to those for the gyrA house-keeping gene were used to determine gene presence or absence . The potential for inversion of the promoter region proceeding the recombinase was assessed by comparison of SZO08560 mRNA transcript levels ( produced when the promoter region is in the forward orientation ) with reverse strand SZO08550 mRNA transcript levels ( produced when the promoter region is inverted ) in SzH70 . SzH70 was grown to log phase in THB with 10% horse sera . A quantitative two-step reverse transcription ( RT ) PCR procedure was used to analyze levels of SZO08560 and reverse strand SZO08550 transcription relative to the housekeeping gene gyrA . RT was performed using the Verso cDNA kit ( Abgene ) . The RT reaction mixture ( 20 µl ) contained 100 ng total RNA , 2 µM gene-specific primer ( ZM474R or ZM476F ) ( Table S5 ) , 500 µM dNTP mix , 1× cDNA synthesis buffer , 1 µl RT enhancer and 1 µl Verso enzyme mix . RT was performed at 50°C for 30 min and terminated by heating to 95°C for 2 min . Quantitative real time PCR ( QPCR ) was performed with a Techne Quantica instrument and data analyzed using Quansoft software ( Techne ) . For the QPCR , 6 µl RT reaction mixture diluted 1/1000 was mixed with 0 . 3 µM forward and reverse primers ( Table S5 ) , and 1× ABsolute QPCR SYBR green mix ( Abgene ) in a total volume of 20 µl and subjected to thermocycling at 95°C for 15 min , followed by 40 cycles of 95°C for 15 s , 55°C for 30 s and 72°C for 30 s . Dissociation curves were analyzed , following a final ramp step from 60°C to 90°C with reads at 0 . 5°C increments , to rule out non-specific amplification . No-template negative controls were included and reverse transcriptase negative controls to confirm the absence of contaminating DNA from RNA samples . Standard curves ( Crossing point vs . log gene copy number ) were generated from genomic DNA for each target gene and used to calculate transcript copy number in cDNA samples . SZO08560 and reverse strand SZO08550 transcript copy numbers were normalized to gyrA reference gene copy number to correct for differences in the amount of starting material . Data was expressed as fold difference in normalized SZO08560 transcript level relative to reverse strand SZO08550 transcript level . Phage particle DNA was purified according to previously published methods [70] . Se4047 was grown to log phase and treated for 3 hours with mitomycin C . Bacteria were centrifuged at 8 , 000×g for 15 minutes and the supernatant was sterilized with a 0 . 45 µm filter ( Millipore ) . The filter-sterilized supernatant was centrifuged at 141 , 000×g for 4 h at 10°C , and the pellet resuspended in 1 ml phage suspension buffer . 0 . 5 ml phage particles were treated with 25 U benzonase ( Novagen ) for 1 h at 37°C and then lysed with 0 . 5% sodium dodecyl sulfate , 10 mM EDTA and 500 µg of proteinase K ( Sigma ) /ml for 1 h at 37°C . Phage DNA was extracted with an equal volume of phenol-chloroform-isoamyl alcohol ( 25∶24∶1 ) ( Sigma ) , followed by an equal volume of chloroform-isoamyl alcohol ( 24∶1 ) ( Sigma ) . Phage DNA was precipitated with 300 mM NaOAc ( pH 4 . 6 ) ( Sigma ) and a 2 . 5-fold volume of ethanol at −20°C overnight , washed with 70% ethanol and suspended in distilled H2O . Prophage induction was detected by PCR with forward and reverse primers ( Table S5 ) that were specific for each recircularized prophage and amplified across the join of prophage ends . Se4047 genomic DNA was used to confirm that the integrated prophage did not generate a PCR product using these primers . PCR products generated from phage particle DNA preparations were purified on QIAquick spin columns ( Qiagen ) and the sequences of both strands of the PCR fragments were determined using an ABI3100 DNA sequencer with BigDye fluorescent terminators and the primers used in the initial PCR amplification to confirm prophage recircularization . The sequence and annotation of the Se4047 and SzH70 genomes have been deposited in the EMBL database under accession numbers FM204883 and FM204884 , respectively . | Streptococci colonize a diverse range of animals and tissues , and this association is normally harmless . Occasionally some strains of streptococci have an increased ability to cause disease that is often associated with a reduction in the ability to colonize and the acquisition of new genes , which enable the strain to inhabit a new niche . S . equi is the causative agent of strangles , one of the most frequently diagnosed and feared infectious diseases of horses , which is believed to have evolved from the closely related and usually harmless S . zooepidemicus . We aim to understand the mechanisms by which S . equi causes disease by studying and comparing the genomes of these different strains . Here we identify specific genes that have been lost and gained by S . equi , which may have directed its transition from colonizer to invader . Several of the novel genes acquired by S . equi have also been identified in strains of the closely related bacterium S . pyogenes that are associated with increased morbidity and mortality in humans . Our research highlights the role of genetic exchange in cross-species bacterial evolution and argues that the evolution of human pathogens cannot be considered in isolation . | [
"Abstract",
"Introduction",
"Results/Discussion",
"Materials",
"and",
"Methods"
] | [
"genetics",
"and",
"genomics/microbial",
"evolution",
"and",
"genomics",
"evolutionary",
"biology/microbial",
"evolution",
"and",
"genomics",
"genetics",
"and",
"genomics/gene",
"discovery",
"microbiology/microbial",
"evolution",
"and",
"genomics",
"genetics",
"and",
"genomics/genome",
"projects",
"infectious",
"diseases/respiratory",
"infections",
"infectious",
"diseases/bacterial",
"infections",
"genetics",
"and",
"genomics/population",
"genetics"
] | 2009 | Genomic Evidence for the Evolution of Streptococcus equi: Host Restriction, Increased Virulence, and Genetic Exchange with Human Pathogens |
ComK transcriptionally controls competence for the uptake of transforming DNA in Bacillus subtilis . Only 10%–20% of the cells in a clonal population are randomly selected for competence . Because ComK activates its own promoter , cells exceeding a threshold amount of ComK trigger a positive feedback loop , transitioning to the competence ON state . The transition rate increases to a maximum during the approach to stationary phase and then decreases , with most cells remaining OFF . The average basal rate of comK transcription increases transiently , defining a window of opportunity for transitions and accounting for the heterogeneity of competent populations . We show that as the concentration of the response regulator Spo0A∼P increases during the entry to stationary phase it first induces comK promoter activity and then represses it by direct binding . Spo0A∼P activates by antagonizing the repressor , Rok . This amplifies an inherent increase in basal level comK promoter activity that takes place during the approach to stationary phase and is a general feature of core promoters , serving to couple the probability of competence transitions to growth rate . Competence transitions are thus regulated by growth rate and temporally controlled by the complex mechanisms that govern the formation of Spo0A∼P . On the level of individual cells , the fate-determining noise for competence is intrinsic to the comK promoter . This overall mechanism has been stochastically simulated and shown to be plausible . Thus , a deterministic mechanism modulates an inherently stochastic process .
Stochastic gene expression during development has received much attention in both bacterial and eukaryotic systems [1] , [2] , [3] . The developmental choices exhibited by the model bacterium Bacillus subtilis have emerged as favorite subjects for the analysis of stochastic decisions [4] , [5] , [6] , [7] , [8] , [9] , [10] , [11] . When faced with the exhaustion of nutrients , individual bacteria may sporulate , become “competent” , or may form multi-cellular communities ( biofilms ) [12] , [13] , [14] , [15] . Remarkably , all of these adaptive responses are expressed heterogeneously in populations . For example , only about 10–20% of the cells in a culture express the proteins needed for the uptake of DNA [16] , [17]; most of the cells therefore remain in the non-competent state . The selection of cells for competence is random , driven by noise in the expression of a transcription factor gene , comK [5] , [6] , [8] . ComK is the proximal regulator of competence gene expression . This protein not only activates transcription of the downstream genes that encode DNA-uptake proteins , but also positively regulates its own promoter , and activation of the resulting positive feedback loop is at the heart of the decision to become competent [18] , [19] , [20] . Cells with more than a threshold amount of ComK induce further expression of comK and as a result can express the downstream genes . Explicit models have been suggested to explain the behavior of this system and experimental evidence has served to test and confirm the general validity of the models , particularly the central roles of positive feedback , a thresholded response to ComK and of noise in comK expression [5] , [6] , [8] , [19] , [20] , [21] . During long-term cultivation , cells have been observed to enter and then exit the competent state , prompting the description of competence as an excitable system [5] . In bulk cultures grown to saturation , the rate of transitions to the competence-ON state increases rapidly as the culture approaches stationary phase and then decreases , approaching zero 1–2 hours later , so that most cells never express competence genes [8] , [11] , [22] . An explanation for this “window of opportunity” was provided by the observation that the basal level of comK mRNA increases during growth and then decreases , reaching a maximum level of 1–2 transcripts per cell near the time of the highest transition rate [8] , [11] . Thus as the average mRNA content ( and therefore the average ComK content ) of the population increases , more cells cross the threshold and become competent . As the rate of transcription then approaches zero and comK transcripts decay , the average ComK content per cell no longer increases , or does so slowly . Cells that were already across the threshold would become competent with high probability and cells below the threshold would tend to remain in the OFF state . The increase and decrease in comK promoter activity ( referred to hereafter as the “uptick” ) was thus assigned the important role of defining the window of opportunity for making competent cells . In other words , a programmed regulatory change would adjust the probability of this developmental transition , modulating a stochastic decision-making process . We show here that the phosphorylated form of the response-regulator protein Spo0A both activates and represses the basal firing rate of PcomK and does so by direct binding to a series of activator and operator sites . Hence , the gradual increase in Spo0A∼P ( OA∼P throughout ) as cells enter stationary phase , which is documented in this report , influences the uptick profile of PcomK activity . Remarkably , 0A∼P acts on PcomK by amplifying a general increase in the rate of transcription also characteristic of synthetic “core” promoters containing only the −35 and −10 RNA polymerase ( RNAPol ) recognition motifs , which takes place as growth slows . This amplification is largely due to the action of 0A∼P as an anti-repressor toward the repressor protein Rok . As predicted by this mechanism , we show that a cell destined for competence must have a level of 0A∼P within a certain permissive range; below and above this range the probability of transition is low . We also show that noise in the level of 0A∼P within the permissive range makes a relatively small contribution to the determination of which cells transition to competence . This choice depends instead on intrinsic cell-to-cell variation in the firing of PcomK . We propose that an extrinsically determined increase in the average firing rate of core promoters and a programmed increase in 0A∼P concentration define the window of opportunity for entrance to the competent state .
Based on measurements in single cells , we previously reported an uptick in the content of comK transcripts and it has been proposed that this variation determines the probability of transition to the competent state [8] , [11] . Figure 1A shows that the uptick is also evident from ensemble measurements , using a fusion of PcomK to the firefly luciferase coding sequence ( luc ) . For this , the cultures were grown in a temperature-controlled plate reader in the presence of luciferin , as previously described [23] , a method characterized by remarkable reproducibility ( Figure S1 ) . It is important to recognize that luciferase activity reports the transcription rate , rather than the cumulative activity of PcomK [23] . As shown in Figure 1A and 1B , and confirming the previous result obtained with fluorescence in situ hybridization , the uptick is independent of comK . Thus , the basal level of comK promoter activity increases in the null-mutant comK strain during growth , reaching a maximum when the growth rate decreases , and thereafter declines . In a comK+ strain , the initial rise is identical , but as the positive feedback is activated and cells transition to competence expression , a massive induction of PcomK takes place , obscuring the decline in the basal level expression that is known to take place in the non-competent cells [8] , [11] . Using the PcomK-luc fusion in a comK null-mutant strain , we adopted a candidate approach to determine factors involved in the uptick , introducing null-mutations in each of the genes known to encode regulatory factors for competence; rok , spo0A , degU , abrB , codY and sinR . Although several of these mutations slightly modified the amplitude of the uptick ( not shown ) , only spo0A and rok exhibited major effects . In a spo0A null background ( Figure 1B and 1C ) , the amplitude of the uptick was decreased about 9-fold , consistent with the observation that a spo0A loss-of-function mutant does not express competence genes [24] . The negative effect of the spo0A null mutant was not reversed by inactivation of abrB , showing that it is not due to over-expression of this repressor ( not shown ) . When plotted to equalize the peaks of the spo0A+ and Δspo0A strains , the increasing portions of the uptick in the two strains are closely similar ( Figure 2A ) . In both strains the downward portion begins at the same time , but that of the spo0A+ strain is more abrupt ( Figure 2A ) . It appears that Spo0A amplifies an underlying increase in the transcription rate of comK and also contributes to the relatively sharp decrease in the transcription rate that normally takes place . Thus , we predict that OA∼P may exert both positive and negative effects on PcomK , depending on its time of action . All of the known activities of Spo0A are dependent on its phosphorylation by a phosphorelay , in which several kinases feed phosphoryl groups to a cascade of three proteins , the last of which is Spo0A [25] . The activity of Spo0A just described is no exception; inactivation of a phosphorelay gene ( spo0F ) and of two kinases ( kinA and kinB ) decreases the uptick amplitude ( Figure S2 ) . As previously reported , in the absence of Rok , a repressor of PcomK [26] , the amplitude of the uptick was enhanced markedly but the timing of its onset was not altered [8] , [26] . ( Compare the ΔcomK and ΔcomK Δrok strains in Figure 1C . ) The increased amplitude is consistent with the published observation that the fraction of competent cells increases to 70–80% in a Δrok comK+ strain , while the time of transition to competence is not changed [26] . Because the average basal level of comK expression is increased in the rok strain , more cells exceed the threshold for competence . Importantly , when rok was inactivated in the ΔcomK Δspo0A background , a dramatic recovery of PcomK-luc activity was observed , to about 60% of the level of the ΔcomK Δrok spo0A+ strain ( Figure 1C ) . This recovery suggests that 0A∼P may act at least in part by antagonizing the action of Rok . These changes in PcomK activity were not due to an effect of spo0A inactivation on the amount of Rok protein ( Figure S3 ) and we show below that they are explained instead by direct interactions of Rok and 0A∼P with the promoter of comK . In the two Δrok strains shown in Figure 1C the decreasing segment of the expression profiles was altered by the appearance of a pronounced shoulder . The evidence presented so far suggests that Rok and 0A∼P contribute to the normally abrupt downturn in transcription rate from PcomK as well as acting respectively as repressor and activator of PcomK earlier in the development of competence . We wondered if the changes in PcomK expression observed even in the Δspo0A strain ( Figure 2A ) were specific to the comK promoter or were more global in nature . To answer this question we tested a synthetic SigA-dependent promoter derived from random sequences , except for the presence of canonical −35 and −10 motifs ( PsynthA ) ( Figure S4 ) . This promoter was fused to the luc coding sequence , preceded by the ribosomal binding site from the spoVG gene . Figure 2B shows that PsynthA exhibited an expression profile very similar to that of PcomK in the ΔcomK Δrok Δspo0A background . A similar result was obtained using a previously described [23] fusion of luc to a stripped-down promoter consisting of the −35 and −10 sequences of the spo0A vegetative ( SigA-dependent ) promoter ( not shown ) . We have also constructed a synthetic promoter carrying Sigma H motifs ( PsynthH ) and fused this to luc . This promoter also showed a pattern like that of the PsynthA promoter ( Figure S4 ) . We conclude that PcomK exhibits a change in activity during the transition to stationary phase , which is a general characteristic of promoter core elements . It also appears that Spo0A and Rok modulate this activity change in the case of PcomK , causing changes in amplitude and collaborating to suppress the shoulder that occurs in the core promoters during the downward response . We suspect that these core promoter changes reflect global metabolic changes during the slowing of growth as cells approach stationary phase , perhaps due to the release of RNAPol from the transcription of stable RNA genes [23] . These effects of rok and spo0A on PcomK activity prompted examination of the sequence surrounding PcomK ( Figure 3 ) . ComK is known to act positively on its own promoter , interacting with ComK boxes located upstream of the −35 SigA recognition motif [27] , while Rok represses PcomK by direct binding [26] . Rok is known to interact at least within an approximately 100 bp region ( Figure 3B ) [28] , which overlaps the ComK boxes and has been shown to bind preferentially to AT-rich sequences [29] . ComK activates PcomK largely as an anti-repressor for Rok , although it does not displace this repressor from the DNA [28] . Inspection of the region surrounding PcomK revealed five sequences resembling the Spo0A binding consensus [30] , which are illustrated in Figure 3A and 3B . The positions of these sequences suggested that three of them ( A1 , A2 and A3 ) functioned as activation sites and that the remaining two ( R1 and R2 ) were sites for repression . None of these putative binding motifs have more than 2 out of 7 differences from the consensus and A1 and 3 have one mismatch each . The presence of the putative 0A∼P binding sites and their locations suggested a role for increasing concentrations of 0A∼P , activating at A1 , 2 and 3 and then repressing at R1 and 2 . Such a mechanism would resemble the action of 0A∼P at the promoter of sinI [31] although in the present case we must also posit a regulatory role for Rok . This model requires that 0A∼P binds directly to PcomK at the sites we have identified . We were encouraged by a previous observation that Spo0A was able to bind to a comK promoter fragment [30] . Accordingly , we next determined whether Spo0A does indeed bind to A1 , 2 , 3 on one hand , and to R1 and R2 on the other . To test the binding of Spo0A to sequences surrounding PcomK , we utilized gel shifts of radiolabeled DNA fragments incubated with purified full length 0A∼P . Figure 4B shows the results of an experiment using a 128 bp fragment ( Figure 4A ) containing A1 , 2 and 3 and an identical experiment with the same fragment in which the putative Spo0A binding sites carried the triple A123 mutations described above ( Figure 3C ) . With the wild type fragment , 50% of the probe was shifted by a Spo0A concentration of about 50 nM . Due to uncertainty concerning the fraction of phosphorylated versus unphosphorylated protein , this is likely to be an overestimate of the nominal KD . The mutant fragment exhibited little or no shift over the protein concentration range used for this experiment . These results strongly suggest that 0A∼P binds to PcomK directly , and that the sequences identified as A1 , 2 and 3 are important for binding , in agreement with the in vivo data in Figure 5 and Figure 6 . A similar experiment ( Figure 4C ) was carried out using a 146 bp fragment ( Figure 4A ) containing R1 and 2 as well as an identical fragment with the mutations that were shown to inactivate these sites in vivo ( Figure 3 , Figure 5E and 5F ) . A shift was noted with the wild type fragment , with a nominal KD of 100–150 nM . Once again the mutant fragment exhibited little or no evidence of binding . To test whether A2 is a binding site for Rok , a gel shift experiment was carried out using the wild type 128 bp probe fragment also used in Figure 4B . Multiple shifted bands were detected ( Figure 4D , upper panel ) with a KD in the low nanomolar range , consistent with results obtained with other promoters [26] . The A2 mutation indicated in Figure 3C was introduced into this probe and the gel shift results obtained with Rok ( Figure 4D , lower panel ) permit two conclusions . First , the nominal KD is shifted from about 15 nM to about 50 nM , and two shifted bands , indicated by arrows in the figure are missing with the mutant probe consistent with a high affinity interaction of Rok with A2 . Second , even with A2 inactivated , residual lower affinity interaction of Rok to this probe remains , suggesting that Rok interacts with additional nearby sites . As a first test of the roles of the putative OA∼P binding sites , we introduced mutations ( Figure 3C ) and determined their effects on PcomK-luc expression in a comK background . Because R1 overlapped the −35 motif we changed this motif to the consensus TTGACA to avoid inactivating the promoter , leaving only three out of seven bases of the −35 consensus . The other mutant Spo0A boxes retained two out of seven consensus bases . Inactivation of either R1 or R2 introduced prominent shoulders in the declining portion of the uptick profile , without changing the timing of the increasing segment ( Figure 5A and 5B ) , supporting a role for these sequences in closing the window of opportunity . Although the amplitude of the R1 shoulder is greater than that of R2 , the timings of the increases and decreases in the two profiles are similar . These profiles also resemble those of the various Δrok and Δspo0A constructs and of the PsynthA-luc and PsynthH fusions ( Figure 1C , Figure 2A and 2B , Figure S4 ) . Figure S5 shows a comparison of the declining portions of the R1 , R2 , PsynthA and wild type promoters and the effects of the Δrok , Δspo0A and combined Δrok and Δspo0A mutations , with their peak values normalized . The roughly similar overall shapes of these curves are consistent with the predicted role for 0A∼P as a repressor that interacts with R1 and 2 , collaborating with Rok in suppressing a continuation of PcomK activity . The observation that inactivation of either spo0A or rok gives a similar slow decrease in the rate of comK transcription and that these rates resemble that when both spo0A and rok are inactivated , suggests strongly that the two proteins work together in repressing comK transcription . Inactivation of either A1 or 3 reduced the uptick to the level of the spo0A null-mutant , in support of the hypothesis that these are activating sites for interaction with 0A∼P ( Figure 5C and 5D ) . Unexpectedly , inactivation of A2 had the opposite effect , increasing expression dramatically , nearly to the level of the rok mutant ( compare Figure 6A and 6B ) . The strain with both the Δrok and A2 mutation exhibited an amplitude equal to that of the rok mutation alone ( compare Figure 6A and 6B ) . These results suggested that A2 might overlap a binding site for Rok , and that 0A∼P may act at this sequence , competing for Rok binding or otherwise interfering with Rok activity . In fact , in the sequence covering A2 and extending between the centers of A1 and 3 , 44 out of 54 bases are either A or T ( Figure 3 ) , and Rok has been found to bind preferentially to AT-rich DNA [29] . When A1 , 2 and 3 were mutated simultaneously , the maximum expression decreased by about half from the level of the A2 mutant alone ( compare Figure 6A and Figure 7A ) showing that A1 and 3 retained an activating role even when A2 was inactivated . When rok was eliminated and A1 , 2 and 3 were also mutated , the amplitude of the profile increased slightly compared to the A123 triple mutant in the rok+ background ( Figure 7B ) , much less than the nearly five-fold increase caused by inactivation of rok in the wild type promoter background ( Figure 1C ) . It is likely that our mutations have not completely eliminated Rok binding . This conclusion is supported by the demonstration that Rok binding with lowered affinity is retained nearby when A2 is inactivated ( Figure 4D ) . When a spo0A knockout was introduced into the A123 strain , a small decrease in amplitude ( ∼30% , compared to a nine-fold decrease with the wild type promoter ( Figure 1B ) ) was observed ( Figure 7B ) , supporting the conclusion that the major role of 0A∼P is to antagonize Rok . The working model suggested by the above data requires that both the increase and decrease of comK basal transcription are largely determined by a continual rise in the amount of OA∼P . We have shown that the rate of spo0A transcription increases and is then maintained throughout the time of the uptick , with a pattern similar to that of core promoters e . g . PsynthA , suggesting that this continual rise does occur ( Figure S6 ) . In addition , the transcription of a luc fusion to the OA∼P-dependent sdp promoter increases throughout the uptick ( not shown ) , confirming that the amount of OA∼P in the cell increases . The data suggest that during the upward segment of the PcomK uptick , 0A∼P amplifies an inherent rise in PcomK activity by binding to A1 , 2 and 3 , thereby interfering with the activity of Rok . Three pieces of evidence support this role of OA-P as an anti-repressor . First , the inactivation of A2 , identified from its sequence as an OA∼P binding site , imparts a Rok-like phenotype . Second , A2 and its surrounding sequence is AT-rich , as expected for a Rok binding site and third , in a Δrok background , the Δspo0A phenotype is essentially lost . The inherent rise that is amplified by 0A∼P is a general one , reflecting some aspect of cell physiology that occurs as growth slows . In addition to its activity as an anti-repressor of Rok , there must be another , relatively minor component of 0A∼P activation . This is apparent from a comparison of the uptick in a ΔcomK Δrok Δspo0A strain with a ΔcomK Δrok spo0A+ strain ( Figure 1C ) . This minor component may be due to interference with the activity of CodY , which plays a relatively small role in repression at PcomK [28] , [32] ( and results not shown ) . It is also possible that 0A∼P acts as a classical activator , contacting RNAPol . The decrease in PsynthA expression begins at the same time as in PcomK , but is less abrupt and possesses a prominent shoulder ( compare Figure 1B with Figure 2B ) . When rok alone or both spo0A and rok were inactivated , a similar shoulder was observed in the decline of PcomK expression ( Figure 1C , Figure 2B ) and the same was true with the R1 and R2 mutants ( Figure 5A and 5B ) . The data suggest that increasing concentrations of 0A∼P interact with R1 and R2 , reducing expression and that repression by Rok aids in shutting down PcomK expression . Our working model suggests that the uptick will occur within an intermediate range of 0A∼P concentration . As a test of this prediction , we wished to measure the fluorescence intensities in individual cells of a comK+ strain co-expressing fluorescent protein fusions to promoters that report respectively 0A∼P concentration and PcomK activity . To interpret these data , it was first necessary to determine the level of PcomK expression above which cells can be classified as competent . We therefore carried out a calibration experiment using a PcomK-cfp fusion co-expressed with the ComK-dependent promoter PcomG , fused to yfp , collecting data at 1 . 5 and 4 hours after the onset of stationary phase ( Figure S7 ) . Before 1 . 5 hours , the fluorescent signals are too weak to permit accurate data collection . The average expression of both markers increased between the two time points . In Figure S7B , most of the cells cluster with low levels of both YFP and CFP . Those with higher amounts of PcomG-yfp expression have more than 50 units of CFP fluorescence . These cells comprise 13 . 3% of the total population , in good agreement with the range of values ( 10–20% ) reported in the literature for the fraction of cells that become competent [8] , [11] , [20] , [33] . This value was transferred to Figure S7A , in which a box is shown surrounding cells which fall within the upper 13 . 3% of CFP fluorescence . It is satisfying that the lower edge of this box includes all of the cells with elevated expression of comK-cfp , showing that the transitions had reached a limit at the earlier time , as expected [8] , [11] . This lower limit of PcomK-cfp expression at the earlier time point corresponded to a value ( in arbitrary units ) of 36 . This value was then used for further experiments , to determine which cells had expressed sufficient comK at 1 . 5 hours after the onset of stationary phase to eventually transition to competence . ( For comparison , the background signal due to auto-fluorescence in cells not encoding cfp is 12–16 ) . One other feature of Figure S7 is noteworthy . Although the high cfp values tend to occur in cells with high yfp values , there is considerable scatter . At a given level of PcomK-cfp expression , many values for PcomG expression are possible , and vice versa . This may reflect the time of transition , so that cells that became competent earlier have more YFP . There may also be significant noise in the expression of PcomG . sdp encodes a toxin , required for cannibalism , that is produced by cells that possess low to intermediate levels of OA∼P [34] . Psdp is activated by 0A∼P and in a population of cells provides a sensitive reporter of the average concentration of this molecule [35] . Figure 8A presents data from cells that co-express PcomK-cfp and Psdp-yfp . These data were collected 1 . 5 hours after the onset of stationary phase and may therefore be compared with the calibration results in Figure S7A . In Figure 8A , a green box encloses cells that exceed the threshold value of 36 for PcomK-cfp and these were found to comprise 12 . 6% of the total population . This value is close to the value of 13 . 3% determined in Figure S7A . Inspection of the scatter plot in Figure 8A suggests that above a YFP fluorescence level approximated by the vertical dashed line there are very few competent cells . In panel A there are a total of 11 , 238 cells represented , of which 1 , 419 have been classified as competent ( within the green box ) . Of these cells , 48 are to the right of the dashed line , and these comprise 2 . 9% of all the cells to the right of this line . The equivalent percentage of competent cells with Psdp expression lower than this limit ( to the left of the dashed line ) is 14 . 4 . These proportions differ with P<0 . 001 , based on a two-tailed t-test . This decrease in the probability of transitions with high 0A∼P is predicted by the model for the 0A∼P dependence of PcomK expression presented above . In this experiment , two types of cells may possess a high level of YFP without transitioning to competence . Some may have experienced a rapid rise in the concentration of 0A∼P , which reached an inhibitory level before PcomK could be expressed . Others may have experienced a gradual increase in 0A∼P but simply by chance did not transition to high PcomK expression . The histogram in panel A shows the distribution of competent cell numbers , as a percent of the total number of cells within each “slice” of YFP fluorescence . This distribution confirms that competence transitions occur within an intermediate range of Psdp expression and therefore within an intermediate concentration range of 0A∼P . It also suggests that there is an optimal 0A∼P concentration that maximizes the probability of competence . Based on the following argument , it was of interest to compare these results with an analogous experiment in which PcomK-cfp was co-expressed with Pspo0A-yfp ( Figure 8B ) . A priori , it is not known whether Pspo0A-yfp expression would accurately reflect the level of 0A∼P in the cell because of two uncertainties . First , at a given time all the Spo0A may not be phosphorylated . Second , if the cell-to-cell variation in transcription of spo0A were dominated by intrinsic noise , there would be little correlation between the YFP fluorescence of a given cell derived from Pspo0A-yfp and the amount of Spo0A protein , encoded by the wild type spo0A gene [36] . Comparison of the two panels of Figure 8 shed some light on these uncertainties and therefore has implications for the kinetics of Spo0A phosphorylation and the nature of noise in the amount of this molecule . As in panel A , the data in panel B shows a paucity of competent cells above a level of YFP approximated by the vertical dashed line . A calculation comparable to the one described above confirms that this paucity is highly significant ( P<0 . 001 ) . This comparison strongly suggests that the variation from cell-to-cell in the rate of expression from Pspo0A-yfp is dominated by extrinsic noise , so that expression from the wild type and Pspo0A-yfp promoters in each cell is correlated . If there were a major contribution from intrinsic noise in the firing of Pspo0A , there would be little relationship between the expression from Pspo0A and from PcomK and the points in panel B would therefore extend over the entire range of Pspo0A-yfp values . We have observed that the pattern of luciferase expression from Pspo0A is quite similar to that of PsynthA ( Figure S5 ) , suggesting that a global change in the physiology of these transition state cells may be responsible for changes in the average amount of spo0A expression and perhaps also for extrinsic noise in expression from Pspo0A . A relative unimportance of intrinsic noise would be expected if Spo0A were abundant in these cells [37] and in fact it has been estimated that there are about 2000 molecules of this protein per cell during growth [38] . Further , the similarity in the patterns of PcomK-cfp expression in the two panels , including in the respective histograms , is consistent with the notion that the phosphorelay is finely tuned to rapidly phosphorylate newly synthesized Spo0A as it becomes available . As a whole , these results support the conclusion that 0A∼P plays both positive and negative roles in determining the basal level of comK expression and hence the probability of transitions to competence . As noted above , the histograms shown in Figure 8A and 8B suggest that there is an optimum level of 0A∼P for competence . However , at each level of 0A∼P , most cells do not transition to competence and there must be another source of noise that determines whether a given cell becomes competent . We will return to this point below . To test the plausibility of OA∼P's dual role as a positive and negative regulator of the comK promoter , we simulated the network with Gillespie's stochastic simulation algorithm [39] with our primary output being expression from the comK promoter over time . Since our objective was to verify that Rok and Spo0A alone could define an uptick in comK expression , we only incorporated the binding sites for Spo0A and Rok described above . In addition , we developed a simple model to account for the global decline in transcription following entry into the stationary phase of the growth cycle ( see Text S1 for details of the model , including all parameter values ( Table S3 ) and the reactions ( Table S4 ) ) . With the appropriate parameters , we found that our model captured the temporal dynamics of the comK promoter and qualitatively captured the relative amplitudes of the various null-mutant strains ( Figure 9A , analog to Figure 1C ) . By normalizing the amplitudes of the uptick in each strain , one can see the sharpened decay in the ΔcomK Δrok strain ( Figure 9B , analog of Figure S5 ) . Our simulations suggest the possibility of an additional role of OA∼P as a classical activator that recruits RNA polymerase to the promoter . If OA∼P only activated comK expression as an antirepressor of Rok , then the promoter with OA∼P bound at sites A123 would have the same probability of recruiting polymerase as the bare promoter without any Rok . In our simulation output , however , we observed that the ΔcomK Δrok and the ΔcomK Δrok Δspo0A strains yielded identical expression of comK unless OA∼P enhanced the transcription rate when bound at sites A123 . Given that null-mutant strains of other regulatory factors of competence showed no major change in the uptick amplitude when measured by the luciferase assay , our simulation favors the additional role of OA∼P as a direct activator of comK . Using our simulations , we addressed the question of whether fluctuations in OA could transmit to fluctuations in the expression of comK , thus potentially determining which cells become competent and which do not . We incorporated two independent comK promoters with identical binding affinities and observed that the number of transcripts from the two promoters was largely uncorrelated ( the correlation coefficient ranged from 0 . 06 to 0 . 22 ) at various time points in the simulation ( Table S5 ) . This supports the view that variability in Spo0A is not a major determinant in whether or not a cell becomes competent , but rather that comK noise is the major factor as suggested previously [8] . Instead , the concentration of 0A∼P serves only to determine the window by shaping the probability of whether or not cells will become competent , without explicitly dictating cell fate .
The probability that a cell will enter the competent state increases and then decreases as a culture departs from exponential growth . The present work identifies two temporally variable factors that determine this transient uptick ( Figure 10 ) . First , we have discovered a previously unsuspected increase in the basal firing rate of the PcomK promoter as cells approach stationary phase , which seems to be characteristic of core promoter elements in general , even one that depends on a minor sigma factor . This increase must be due to factors extrinsic to the “core” promoters themselves and may reflect the metabolic state of the cell , e . g . the content of nucleotide pools , the conformational state of the chromosome or the release of RNAPol from the transcription of stable RNA as growth slows [23] . In the cartoon representation of Figure 10 , we have arbitrarily represented this extrinsic factor as an increase in available RNAPol . In any event , the extrinsic change may serve to couple competence regulation to cellular physiology , particularly to the growth rate , so that as growth slows , the probability of transitions to competence would increase . It is possible that a similar increase in the rate of spo0A transcription ( Figure S6 ) also contributes to other forms of Spo0A-dependent stationary phase expression . The second variable factor is 0A∼P , which binds to A1 , 2 and 3 , amplifying the increase in PcomK activity that is due to the core promoter activity , acting principally by antagonizing the repressing effect of Rok . During the rising portion of the uptick an increasing concentration of 0A∼P antagonizes Rok ( and perhaps CodY as well ) ( Figure 10 ) . In randomly selected cells , this causes the concentration of ComK to approach the threshold for transition to competence , raising the global probability that the promoter will fire . PcomK activity decreases more abruptly than that of the core promoter ( Figure 2B ) , because as the concentration of 0A∼P rises , it begins to occupy R1 and R2 . Rok contributes to this decrease because it acts as a repressor , helping OA∼P shut off transcription . It is likely that Rok does this by acting at a site in addition to A2 , as represented in Figure 10 , because the downturn in the A2 mutant strain is more abrupt than in the Δrok A2 double mutant ( Figure 6B ) . Thus , 0A∼P allows transitions to occur , but also closes the window of opportunity . Rok modulates this process as a repressor by working against 0A∼P during the upward part of the uptick and by accelerating the down turn , but does not itself vary in amount ( Figure S3 ) . In this model , the behavior of the core promoter and the increase in the concentration of 0A∼P are the primary variable factors that determine the timing of the uptick profile . Stochastic simulations of the model described here successfully reproduced the experimental behavior of the system , including the relative amplitudes and the temporal dynamics of comK basal transcription in the wild type and mutant strains . Furthermore the modeling assumption , based on experimental evidence , that OA∼P and Rok assist one another in effecting repression at PcomK , successfully reproduced the dynamics of repression in single and double rok spo0A mutants ( Figure 1C , Figure 9 , and Figure S5 ) . Although the model described above is plausible , we cannot be certain about the individual roles of A1 , A2 and A3 and it is prudent to consider alternative possibilities . For example , because inactivation of A2 results in an increase in the amplitude of the uptick , it is conceivable that A2 is a repression site for OA∼P binding . As noted above , this is unlikely because inactivation of rok in the A2 mutant background has little effect ( Figure 6A and 6B ) , suggesting strongly that Rok and OA∼P work on the same pathway . A similar conclusion can be drawn from comparison of single and double mutants in rok and spo0A ( Figure 2 ) . Although the inactivation of rok alone increases the amplitude of the uptick , further deletion of spo0A fails to reduce this amplitude to the level of the spo0A single mutant . This also suggests that Rok and OA∼P work on the same pathway . Another complex and unlikely possibility is that OA∼P recruits Rok for binding to A2 , explaining why inactivation of either A2 or Rok have the same effect on the uptick . However , Figure 4D shows that Rok can bind to A2 with high affinity in the absence of OA∼P . The regulation of the comK uptick now joins two other cases of timed regulation dependent on 0A∼P , in which this molecule acts positively and then negatively as its average concentration increases . Spo0A acts to cause a burst in the expression of sinI , a key regulator in the early development of biofilms , which is expressed bimodally [15] , [31] . The regulation of sinI has been explained by differential affinity of 0A∼P for activating and repressing sites [31] . The second case concerns sdp , an operon needed for cannibalism , which encodes a toxic extracellular factor that kills non-producing cells with less 0A∼P , permitting the producing cells access to nutrients and thereby delaying their commitment to sporulation [34] . sdp transcription is activated and then repressed by 0A∼P [35] . However , this regulation differs from than of comK and sinI . sdp is repressed by AbrB and its induction is explained by the extreme sensitivity of the abrB promoter to repression by 0A∼P and also by the activation of abbA transcription . The AbbA protein binds to AbrB , further derepressing sdp [40] . High concentrations of 0A∼P , on the other hand , repress sdp transcription by direct binding . Thus , comK , sinI and sdp are all turned on by 0A∼P at low concentrations and turned off as the amount of 0A∼P increases , with this versatile protein acting by various combinations of activation , repression and anti-repression . comK is the only known target gene for which 0A∼P acts as an anti-repressor . Competence thus joins sporulation , cannibalism and biofilm formation as subject to temporal regulation by Spo0A∼P . As B . subtilis enters its developmental pathways , it uses the gradual accumulation of 0A∼P as a clock , to orchestrate the timing of gene expression . Figure 8A shows that competence develops within an intermediate range of Psdp-yfp concentrations . The few cells with elevated CFP at the low end of the YFP distribution is consistent with the observation that the uptick is amplified by Spo0A∼P , although the number of cells in this region of the scatter plot is too low to establish the statistical significance of this observation . Nevertheless , it is known that spo0A mutant cells are not competent [24] and these cells are likely to be below the threshold for competence transitions . Cells at the high end of the YFP distribution likewise have a low probability of being competent because their transition windows have closed . This upper limit is due to repression exerted at R1 and R2 , as shown here . Within the intermediate range of YFP concentrations , corresponding to an intermediate range of Spo0A∼P , conditions are permissive for transitions to occur . Clearly at each concentration of Spo0A∼P within this permissive range , indicated by expression of Psdp-yfp , a wide range of ComK-CFP concentrations can exist and most cells fail to become competent ( Figure 8A ) . Likewise , at each level of PcomK-cfp expression above the threshold value for competence , a range of YFP values are possible . It follows from these results that cell-to-cell variation in the content of Spo0A∼P alone does not determine cell fate . In agreement with this conclusion , our previous evidence suggested that the major source of noise for fate determination is the firing rate of PcomK [8] . The data in Figure 8 are consistent with that result , as is the uptick model that we advanced above . Thus , we suggest that the noise in transcription from PcomK determines cell fate and that 0A∼P amplifies PcomK expression , permitting about 10–20% of the cells to exceed the transition threshold for amplification of comK expression by auto-induction . 0A∼P then shuts down transcription when it reaches higher concentrations , driving the transition rate toward zero . Our earlier experiments suggested that the PcomK noise is intrinsic , because the numbers of transcripts derived from two comK promoters in the same cell were poorly correlated at the peak of the uptick [8] , when there is about one comK transcript in the average cell [8] and therefore the average number of ComK protein molecules is likely to be low . Under these conditions we would expect intrinsic noise in comK expression to dominate cell-to-cell variation [37] and to be the major factor selecting cells for the competent fate . The simulation results described above support this conclusion , shown by the poor correlation observed when the outputs of independent comK promoters were compared ( Table S5 ) . Although noise in comK is the major determinant , 0A∼P acts as a temporal gate , modulating the probability that random fluctuations in comK transcription will lead to competence . It has been well established that a gradual increase in the concentration of OA∼P governs development in B . subtilis . Although competence was known to require spo0A , the role of OA∼P in this pathway was imperfectly understood . It now appears that the probability of transition to the competent state is determined by the programmed increase in OA∼P , as are the development of biofilms , cannibalism and spores . OA∼P thus plays a general role in the temporal control of development and competence must be subject to the complex signaling that governs the synthesis of this molecule , involving multiple kinases , phosphatases and extracellular peptides . Despite this commonality , competence differs from sporulation , cannibalism and biofilm formation in an important respect . Asynchrony during the gradual post-exponential increase in OA∼P concentration is most likely the major determinant of which cell is selected for these last three fates . In contrast , it is noise in comK , a dedicated gene , that governs the choice of cells for competence , although OA∼P helps to set the temporal window of opportunity .
Bacillus subtilis strains were constructed by transformation into BD630 ( his leu met ) with selection for the appropriate antibiotic resistance marker , and all comparisons were with isogenic strains . For transformation , competent cultures were prepared and incubated in competence medium with transforming DNA ( ∼1 µg/ml ) for 30 min at 37°C [24] . The strains are listed in Table S1 . All the growth experiments were carried out in competence medium [24] . For the detection of luciferase activity , strains were first grown in LB medium to an optical density at 600 nm ( OD600 ) of 2 . Cells were then centrifuged and resuspended in fresh competence medium , adjusting all the cultures to an OD600 of 2 . These pre-cultures were then diluted 20 fold in fresh competence medium and 200 µl was distributed in each of two wells in a 96-well black plate ( Corning ) . 10 µl of luciferin was added to each well to reach a final concentration of 1 . 5 mg/ml ( 4 . 7 mM ) . The cultures were incubated at 37°C with agitation in a PerkinElmer Envision 2104 Multilabel Reader equipped with an enhanced sensitivity photomultiplier for luminometry . The temperature of the clear plastic lid was maintained at 38°C to avoid condensation . Relative Luminescence Units ( RLU ) and OD600 were measured at 1 . 5 min intervals . The data were plotted as RLU/OD versus time from the beginning of growth . A 1 Kb fragment ending with the initiating codon of the gene of interest , and containing the promoter , was amplified by PCR from the B . subtilis chromosome . A single nucleotide was inserted in the primer to restore the correct reading frame . Primers are listed in Table S2 ( PcomK1 and 2 ) . The PCR fragment was cut by KpnI/NcoI in sites present at the extremities of the primers used for the amplification . In parallel , the luciferase gene was cut from plasmid pGL3 ( Promega ) by NcoI/BamH1 digestion . A three-fragment ligation was then carried out between the promoter of interest , the luciferase gene and plasmid pUC18Cm digested with KpnI and BamH1 . The resulting plasmid , pCU18cm-promoter::luc , which cannot replicate autonomously in B . subtilis was used to transform B . subtilis with selection for chloramphenicol resistance , where it integrated by single crossover . This event reconstructed the “normal” regulatory region in front of the fusion and a complete copy of the gene of interest , downstream of the fusion . When necessary , mutations were incorporated in the promoter in front of luc using the ‘Change-It Multiple Mutation Site Directed Mutagenesis’ kit ( USB ) . The primers used in these site-directed mutagenesis constructions are listed in Table S2 . The resulting plasmids were verified by sequencing and then integrated by Campbell-like recombination and the structure of the integration event was verified by sequencing a relevant PCR fragment from the chromosome . A 1 Kb fragment ending with the initiating codon of the gene of interest , and containing the promoter , was amplified by PCR from the B . subtilis chromosome . Primers are listed in Table S2 ( Pspo0A3 , Pspo0A4 , Psdp1 and Psdp2 ) . The PCR fragment was cut by EcoRI/XhoI at sites present at the extremities of the primers used for the amplification . In parallel , the yfp gene was amplified using the primers yfp8 and yfp9 ( see Table S2 ) . A three-fragment ligation was then carried out between the promoter of interest , the yfp gene and plasmid pUC18Cm digested with EcoRI and BamH1 . The resulting plasmid , pCU18cm-promoter::yfp , which cannot replicate autonomously in B . subtilis , was used to transform B . subtilis with selection for chloramphenicol resistance , where it integrated by single crossover . This event reconstructed the “normal” regulatory region in front of the fusion and a complete copy of the gene of interest , downstream of the fusion . The luc gene was amplified from the plasmid pGL3 ( Promega ) using the primers ‘sigAcore’ or ‘sigHcore’ and ‘luc2’ ( see Table S2 ) . The ‘sigAcore’ or ‘sigHcore’ primers allowed us to add the core promoters and the ribosomal binding site of spoVG upstream of the luc gene . The PCR products were cut using BamH1 and EcoRI and cloned in the plasmid pDR111 digested with the same enzymes . The resulting plasmids , pDR111-PsigAcore-luc or pDR111-PsigHcore-luc , were used to transform B . subtilis where they integrated by double crossover at the amyE locus . To inactivate comK , we replaced it cleanly with a tetracycline resistance cassette without using a vector . We first amplified 1 kb fragments upstream and downstream of the gene . These fragments are each flanked with one restriction site at the junction with respectively the ‘start’ or the ‘stop’ of the gene . In parallel , we amplified the tet cassette flanked with the same two restriction sites . The three fragments were then digested and ligated together and the ligated DNA was purified through a QIAquick column . The desired product , corresponding to ligation of the three fragments , was purified from an agarose gel . The purified band was then amplified by PCR using the outside primers previously used to amplify the upstream and downstream fragments . After further purification on a QIAquick column , the full fragment ( flanking sequences+antibiotic cassette ) was used to transform B . subtilis with selection for tet-resistance yielding a double crossover event between the chromosome and the region of homology and replacing the gene with the antibiotic cassette . Spo0A was purified following a published protocol [41] as modified [42] . Non-radioactive phosphoramidate was synthesized [43] and used to phosphorylate the Spo0A . On a Superdex 200 column the resulting Spo0A appeared to be predominantly in the dimer form . Rok was purified as a C-terminal His6 fusion as previously described [26] . The primers used to amplify the probe DNA fragments by PCR ( OAwt , OA123 , ORwt and OR1&2 ) are listed in Table S2 . The 5′ ends of these probe fragments were labeled using [g-32P]-ATP and T4 polynucleotide kinase . For this , about 500 ng of DNA was incubated in 50 µl final volume , with 10 units of enzyme and with 40 µCi of [g-32P]-ATP ( specific activity of 4500 Ci/mmol ) . For the gel shift experiments , a typical assay mixture contained ( in a final volume of 20 µl ) , 10 mM Tris-HCl , pH 8 . 0 , 50 mM NaCl , 1 mM EDTA , 1 mM dithiothreitol ( DTT ) , 5% ( v/v ) glycerol , 0 . 5 µg of bovine serum albumin , approximately 0 . 5 ng of probe and the indicated concentrations of 0A∼P protein . After 5 min of incubation at 37°C , 10 µl of this mixture was loaded onto a native 5% ( w/v ) polyacrylamide TBE Ready Gel ( Bio-Rad ) and electrophoresed in 0 . 5× TBE buffer for 1 h at 100 V cm−1 . The gels were pre-run for about an hour . Radioactive fragments were detected by autoradiography . Gel shifts with Rok were performed as described previously [26] . From B . subtilis cultures growing in the plate reader in competence medium , samples were taken at 1 . 5 or 4 hours after the entrance to stationary phase . Aliquots ( 200 µl ) of each sample were centrifuged and resuspended in 100 µl of PBS . One µl of each sample was placed on a 1% agarose pad . All images were acquired using Volocity v 5 . 1 ( Perkin Elmer ) and a Nikon 90i fluorescence microscope with filters appropriate for detection of CFP and YFP . Segmentation and measurements of pixel intensities were carried out with the Volocity measurement tools . | Populations of bacterial cells sometimes bifurcate into subpopulations with different patterns of gene expression . The soil bacterium B . subtilis becomes “competent” for the uptake of environmental DNA , thus acquiring new genetic information . About 15% of the cells are chosen for expression of the competence genes by stochastic fluctuations in the transcription of comK . When the concentration of ComK exceeds a critical threshold , it activates its own expression , a molecular switch is thrown , and competence ensues in that cell . Here we ask why all of the cells do not eventually throw the switch . We show that the basal level expression of comK increases and then decreases as nutrients are exhausted , so that the number of cells exceeding the ComK threshold rises and falls , opening and closing a window of opportunity for competence . Two factors responsible for this “uptick” in comK expression are: 1 ) a global increase in transcription as cell division slows , and 2 ) a continual rise in the concentration of the master regulatory protein Spo0A-P , which activates and then represses comK as it accumulates . The global increase transmits growth rate information and the increase in Spo0A∼P encodes multiple signals , including the nutritional , replication , and population density status of the culture . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"systems",
"biology",
"biochemistry",
"developmental",
"biology",
"model",
"organisms",
"genetics",
"biology",
"computational",
"biology",
"microbiology",
"biophysics",
"genetics",
"and",
"genomics"
] | 2012 | Spo0A∼P Imposes a Temporal Gate for the Bimodal Expression of Competence in Bacillus subtilis |
Aedes aegypti ( L . ) ( Diptera: Culicidae ) is a highly invasive mosquito whose global distribution has fluctuated dramatically over the last 100 years . In Australia the distribution of Ae . aegypti once spanned the eastern seaboard , for 3 , 000 km north to south . However , during the 1900s this distribution markedly reduced and the mosquito disappeared from its southern range . Numerous hypotheses have been proffered for this retraction , however quantitative evidence of the mechanisms driving the disappearance are lacking . We examine historical records during the period when Ae . aegypti disappeared from Brisbane , the largest population centre in Queensland , Australia . In particular , we focus on the targeted management of Ae . aegypti by government authorities , that led to local elimination , something rarely observed in large cities . Numerous factors are likely to be responsible including the removal of larval habitat , especially domestic rainwater tanks , in combination with increased mosquito surveillance and regulatory enforcement . This account of historical events as they pertain to the elimination of Ae . aegypti from Brisbane , will inform assessments of the risks posed by recent human responses to climate change and the reintroduction of 300 , 000 rainwater tanks into the State over the past decade .
Dengue fever is the 21st Century’s most important mosquito-borne viral illness , exerting a huge economic and health burden in the tropics and sub-tropics [1] . The incidence of dengue has increased 30 fold over the past five decades and the disease is now estimated to affect up to 390 million people each year [1 , 2] . This increase in prevalence is due to substantial growth in urbanization , trade , international travel , and the spread of its major vectors [3] . Historically , the high infection rate of the disease has exerted a severe toll on development and economic progress through lost productivity and costs of mosquito control activities [3–5] . In Australia , four large dengue epidemics that swept through Queensland and New South Wales between 1897 and 1926 , were responsible for a reported 733 deaths , and affected up to 90% of the urban population in each outbreak [6] . Dengue was first isolated in Australia around 1956 , and before this time all dengue cases were based solely on clinical diagnoses [7 , 8] . More recently , a rise in imported cases of mosquito borne diseases into Australia has increased the risk of local transmission of dengue , chikungunya and Zika viruses in all regions where the primary vector , Aedes aegypti ( L . ) is present [9 , 10] . Aedes aegypti is most abundant in northern and some parts of central Queensland . Populations have recently been discovered in the South East of the State , 170 km north of Brisbane , the state capital [11] . It was in Brisbane that Thomas Bancroft first implicated Ae . aegypti as an agent in dengue transmission in 1906 [12] , and this was later confirmed by Cleland in 1917 [13] . At the start of surveys for Ae . aegypti in 1911 , Brisbane occupied ~77 km2 ( 30 mi2 ) and by 1923 had grown to ~163 km2 ( 63 mi2 ) [14] . Brisbane is the largest urban centre of the State with a population of 1 . 18 million inhabitants [15] , an area of 1 , 338 km2 ( 516 mi2 ) and it lies just outside of the southern margin of Ae . aegypti in eastern Australia ( Fig 1 ) [16] . Aedes aegypti has co-evolved with and feeds predominantly upon humans , and uses the many types of artificial containers found in urban locations for oviposition [20 , 21] . As such , this mosquito species is closely associated with the domestic environment and the many microclimates found there [22] . Eggs are desiccation-resistant and can stay dormant during unfavourable periods [21] . Therefore , presence of the species is not entirely dependent on rainfall for the provision of suitable aquatic larval habitats and its persistence is facilitated by the use of artificial water-holding containers [23 , 24] . The unscreened house design typical of Queensland in the early 1900s represented highly favourable habitat for Ae . aegypti adults , providing unrestricted access to a human blood meals and enhancing disease transmission [25] . Once established , Ae . aegypti disperses readily through the urban environment and suitable habitat is constantly being invaded [26] . In the past , these characteristics have allowed Ae . aegypti populations to persist throughout much of Australia . Accordingly , its historical distribution includes Queensland , the Northern Territory ( NT ) , Western Australia ( WA; including the south ) and New South Wales ( NSW; Fig 1 ) , with unconfirmed reports in Victoria [27 , 28] . This range decreased dramatically in the mid-twentieth century , with the last recorded collections in NSW in 1948 , and WA in 1970 [29] . In the NT the last recorded collection was in 1957 [27] , but there were some temporary establishments in 2004 and 2011 which were subsequently eliminated [30 , 31] . Herein we refer to contemporary term “elimination” as the intentional action by humans to remove Ae . aegypti from a defined geographical area [32] . Historically , this action has been referred to as “eradication” by medical entomologists [33–37] . We refer to the elimination of dengue as the reduction of disease transmission to zero incidence in a defined geographical area [32] . In Brisbane during the early 1900s , Ae . aegypti was highly abundant and government agencies set out to eliminate it based on the control successes ( that targeted larval habitat ) observed during the construction of the Panama Canal [38–44] . At some time after the last outbreak of dengue in 1943 , the mosquito was assumed to have disappeared from Brisbane . There are a number of hypotheses suggested by medical entomologists for the wide-spread geographic disappearance of the species , however there are little quantitative data to support these assertions and whether or not they apply to Brisbane [29 , 45] . Rainwater tanks have historically been a conspicuous feature of the Australian rural and urban landscape . Since their introduction in the mid-to-late 1800s , these large structures provided a reliable source of potable and non-potable water to households and commercial premises and inadvertently acted as key larval habitat for Ae . aegypti [27 , 29 , 43 , 46] . In Brisbane , rainwater tanks were first identified as habitat for mosquitoes during the early 1900s [47 , 48] , and regulations were introduced in 1911 to prevent mosquitoes from exploiting rainwater tanks and dwellings [49] . Permanent water storage containers provide a reliable population source by acting as refugia for mosquitoes during times of sub-optimal climatic periods such as drought or extreme seasonal temperatures [23 , 27 , 50 , 51] . Recent modelling suggests the presence of rainwater tanks was an important factor in the historical distribution of Ae . aegypti by potentially buffering the effects of climatic extremes on biological processes at geographical range limits [52] . Despite a keen interest in controlling mosquito related diseases in tropical and subtropical countries around the world , there are few documented examples of elimination and little understanding of the factors that have led to localized Ae . aegypti extinction [53–55] . Here we review data collected from a range of historical sources and evaluate the role of anti-mosquito regulations , surveillance and the removal of rainwater tank infrastructure in the elimination of Ae . aegypti from Brisbane .
We acquired historical records compiled from a range of scientific , government and media sources describing anti-mosquito legislation , mosquito surveillance , regulation enforcement and the presence of dengue and Ae . aegypti in Brisbane . We then asked a series of questions . Firstly , what evidence is there to test the hypothesis that Ae . aegypti is still present in Brisbane ? Secondly , if local elimination occurred , what quantifiable evidence was there to test the hypotheses that the removal of rainwater tanks contributed to the elimination of the species ? Finally , what were the drivers that led to the removal of rainwater tank infrastructure ? To answer these questions , we created a time-line of events and linked direct observations recorded by government surveillance and regulatory enforcement . Scientific publications and annual council and state government reports provide quantitative and subjective measures of a number of factors that potentially affected the historical presence of Ae . aegypti in Brisbane . We collected data from entomological surveys that estimated the prevalence of rainwater tanks on premises and incidence of mosquitoes within these tanks . From the first surveys in 1911 , anti-mosquito policy deemed rainwater tanks non-compliant if they allowed the ingress and egress of mosquitoes [49] . Through enforcement of this legislation , residents found to have larval habitat on their property ( including non-compliant rainwater tanks ) were warned that they were in violation of mosquito regulations , and given a “notice of breach” . Residents were asked to comply within two months , after which patrol members would return and re-survey the property [56] . If the return survey was again in breach of regulations , the resident could be in default , fined and prosecuted . We only included initial surveys ( and excluded re-surveys ) in estimations of total dwellings surveyed and for prevalence and compliance rates of rainwater tanks as a reflection of conditions at that time point . Survey results allowed us to identify temporal trends in the outcomes of entomological inspections; prevalence of tanks on properties , rates of rainwater tank compliance with regulations , mosquito presence in tanks , the number of notices given for a breach of mosquito regulations and the number of notices complied with ( S1 and S2 Tables ) . This is the first time those data have been collated to demonstrate the role of government in the elimination of Ae . aegypti from Brisbane . All Ae . aegypti ‘sightings’ were based on scientific or government publications where identification to species level was done by entomologists ( S2 List ) . To test the hypothesis that Ae . aegypti was still extant in Brisbane over the period from 1887 until 2016 , we applied the method of Jaric and Ebenhard [57] . Briefly , this method extends the work of Solow [58] who provided an equation for inferring extinction based on sighting records over time . Here sightings during the observation period are arranged from first to last and used to express the probability of species presence in relation to the number of time units in which the species was recorded [58] . Authors using this method use P = 0 . 05 as the probability threshold , below which the species can be regarded as extinct [47] . Next , we examine the evidence for whether the mosquito population is extinct by first estimating the probability , p , that mosquitoes are observed each year ( based on 27 sightings over the first 69 years ) , and secondly , estimating the probability that there would be 59 subsequent years of no sightings given the estimated value of p . The probability of observing mosquitoes from one year to the next , p , was estimated as p^ using maximum-likelihood , and a 99% confidence interval , [pl , pu] , was created by inverting a likelihood ratio test-statistic . The confidence interval includes all the values of p for which l ( p;x ) ≥l ( p^;x ) −3 . 317 , where l ( p;x ) , is the log-likelihood function evaluated with a yearly sighting probability of p , x denotes the observed historical data , and 3 . 317 is half the 99th percentile of a χ2 random variable with one degree of freedom . Finally , we use the optimal linear estimation ( OLE ) function from the R package sExtinct ( ) to estimate the date of extinction , with 95% confidence intervals . The OLE method infers time to extinction from a temporal distribution of species sightings [59] . All models and simulation code were created and run in R v3 . 3 . 2 [60] . We estimated rates of dengue during historical outbreaks in Brisbane using minimum infection rates documented in the literature during epidemics ( 70% ) , and population data obtained from government statistics [6 , 13 , 61–63] . Deaths associated with dengue during epidemics were taken from historical medical literature and government population statistics [6 , 13 , 61] . As case data were not technically confirmed , it is possible that some were outbreaks may have been undiagnosed cases of chikungunya or other viral diseases [6 , 64 , 65] . Brisbane was first declared a city in 1902 and by 1911 contained 26 , 645 private dwellings [66] . In 1925 , the city amalgamated with a number of other local authorities to become Greater Brisbane and expanded from 43 , 935 dwellings in 1921 to 68 , 096 by 1933 [67 , 68] . We used government statistics describing the number of dwellings within the council boundary of Brisbane annually from 1895 until 1971 to estimate the total number of tanks ( proportion of dwellings in surveys that had a tank present multiplied by total dwellings in Brisbane for relevant year ) , total non-compliant tanks ( proportion of non-compliant tanks surveyed multiplied by total estimated number of tanks ) , the number of non-compliant tanks with mosquitoes present ( estimated non-compliant tanks multiplied by proportion of tanks with mosquitoes present in surveys ) and the proportion of non-compliant tanks to total dwellings ( estimated total non-compliant tanks divided by total dwellings ) [61] .
Noting that historic records refer to Ae . aegypti by multiple names ( Culex bancroftii , Stegomyia fasciata , Aedes argenteus and Stegomyia calopus; see [69] ) , the first published record of Ae . aegypti in Brisbane was in 1887 [70] and we recorded a total of thirty-seven references from historical records ( Table 1 , S2 List ) . Entomological , house-to-house surveys for the species did not commence until anti-mosquito regulations were introduced in 1911 and , at this time , Ae . aegypti was highly prevalent across the city [14 , 71] . By 1932 , mosquito control was decentralized to local health inspectors and little species-specific data was recorded in house-to-house surveys [56] . A dengue outbreak during 1943 re-intensified mosquito surveys with identification to species ( Table 1 ) [72] . However , once the Second World War was over the role of mosquito surveillance was , again , taken over by generalist local government officers , and council reports after 1947 no longer identified or recorded mosquito presence to species level classification . Subsequent recordings of Ae . aegypti were made by Elizabeth Marks who documented distributions from 1957 until 1981 , surveying the last accounts of larval Ae . aegypti in Brisbane in 1944 [73] , 1948 [74] and 1957 ( Table 1 ) [75] . When inferring extinction likelihood , the total length of the observation period ( T ) was 128 years ( 1887 until 2016 ) , the total number of years where Ae . aegypti was observed ( n = 27 ) , and the period between the last positive survey and last observation was 59 years ( 1957 until 2016 ) . Results suggest the probability that the species is still present based on the average frequency of sightings was calculated as P = 0 . 042 . This value was below our threshold level ( P = 0 . 05 ) and suggests the population does not exist . Using the maximum-likelihood and inverted ratio test-statistic approach , we estimated p^=0 . 391 and assuming the true value of p^ is in the 99% range , we estimate [pl , pu] = [0 . 257 , 0 . 546] . The probability of observing 59 consecutive years with no sightings were estimated from the probability mass function of the binomial distribution as 2 . 45E-8 and 5 . 83E-21 respectively . Therefore , based on the inferred sighting probabilities , there is very little evidence to suggest that the mosquito population is extant given the 59 consecutive years of absence . Finally , using OLE we estimate that from 37 total observations in Brisbane , the species likely went extinct in 1961 ( upper CI = 1974 , lower CI = 1957 ) . A series of dengue outbreaks in Brisbane during the late 19th and early 20th century took a severe toll on the population of Brisbane and its economy [6] . Deaths due to dengue peaked in 1905 and infection rates of 70–90% were reported during all epidemics across the city and until 1926 ( Table 1 ) [13 , 63 , 76 , 77] . The last outbreak in Brisbane during 1942/43 resulted in 646 cases [72] and was precipitated by epidemics occurring in northern Queensland [78] . The last recorded , locally acquired case of dengue in Brisbane was noted in 1948 [74] . Shortages in reticulated water in Brisbane during the 1800s and early 1900s meant rainwater tanks provided the most reliable source of potable drinking water for use around the home [79] . The mean number of water tanks declined from over 1 per dwelling in 1912 to less than 1 in 10 in 1971 ( Fig 2 ) . House-to-house surveys were conducted systematically after the creation of the Brisbane Entomological Section in 1926 , with the city divided up into 88 blocks and house-to-house patrols responsible for surveying a number of blocks per year [42] . During the severe dengue epidemic years from 1897 to 1927 , the mean number of rainwater tanks per dwelling were at the highest recorded ( Fig 2; mean ± SD = 0 . 94 ± 0 . 3 ) . Following the introduction of a large water reservoir to supply Brisbane in 1954 , the mean number of rainwater tanks per dwelling consistently fell below 0 . 50 tanks . Based on the number of dwellings within the city area , we estimated the total number of rainwater tanks at 39 , 513 ( 95% CI ± 0 ) in 1912 , before reaching a maximum of 70 , 794 ( 95% CI ± 515 ) in 1950 , and declining to 24 , 647 ( 95% CI ± 935 ) by 1971 ( Fig 2 ) . The total number of house-to-house surveys to monitor compliance and prevalence of mosquitoes per year by entomologists and inspectors has fluctuated greatly ( Fig 3 ) . The period after the Second World War is of most interest , because this was when the City Council focused the most effort on the elimination of Ae . aegypti [80] . During this period , there was a large increase in annual house-to-house surveys , from 11 , 158 in 1948 , to 91 , 127 in 1964 . The number of surveys decreased after 1964 with the mean dropping to 37 , 436 ( 95% CI ± 8 , 041 ) per year until 1989 . Surveys no longer recorded entomological data after 1971 , although inspectors continued to give notices of breaches in regulations until surveys were discontinued in 1989 . The original anti-mosquito regulations defined mosquitoes as noxious , defined what constituted a breach in regulations , and prescribed ways that the resident must prevent mosquito presence [49] . City Ordinances introduced in 1921 [81] and amended in 1933 [82] , were implemented to allow the enforcement of regulations by local authorities and to set penalties for a breach in these regulations ( for examples of breaches see Table 2 ) [83] . The system of notices provided data on breaches with regulations “notices complied with” , and provided measurement of the success of surveillance programs . Compliance with anti-mosquito regulations typically involved the resident sealing or removing rainwater tanks that were at risk of breeding mosquitoes and removing or treating other potential containers . Between 1913 and 1954 many surveyed dwellings were in breach of regulations ( Fig 4; squares; mean ± SD = 0 . 27 ± 0 . 18 ) . After that time however , the proportion of surveys resulting in a notice for a breach in regulations dropped ( Fig 4; squares; 0 . 05 ± 0 . 05 ) , and reached 0 . 03 by 1974 . Compliance with notices was relatively low between 1934 and the end of the Second World War ( Fig 4; triangles; 0 . 80 ± 0 . 09 ) but during the 1950s and 1960s ( the period when Ae . aegypti disappeared ) most notices resulted in compliance ( Fig 4; triangles; 0 . 95 ± 0 . 04 ) . Reports by government entomologists provided the first qualified evidence that Ae . aegypti was present in rainwater tanks from 1912 [71 , 87] . Surveys in 1923 and 1927 provide further evidence for the presence of Ae . aegypti in rainwater tanks [14 , 43] . Using survey data , we estimated the total number of non-compliant tanks in Brisbane ( Fig 5 ) . When regulations were first introduced in 1911 , the majority of tanks were non-compliant [71] and we estimate a total of 39 , 341 ( 95% CI ± 182 ) non-compliant rainwater tanks in Brisbane at this time ( Fig 5 ) . This was followed by a period of steady decline to 1940 when we estimate 11 , 900 ( 95% CI ± 293 ) non-compliant tanks were present ( Fig 5 ) . A marked increase in the estimated number of non-compliant tanks was observed after the Second World War ( with a maximum of 33 , 931; 95% CI ± 788 ) , due to shortages in labour and materials for repairs ( Fig 5 ) [88] . A rapid decline in the estimated number of non-compliant tanks after 1955 returned to the trends observed before the war , and by 1971 there were only an estimated 4 , 627 ( 95% CI ± 324 ) non-compliant tanks left in Brisbane ( Fig 5 ) . Early entomological surveys recorded the number of non-compliant rainwater tanks that contained Ae . aegypti [71] . The highest proportion containing this species was observed during years when dengue epidemics were most severe ( 0 . 37: 679/1 , 832 , SE ± 0 . 011; Fig 6 ) . After the large dengue epidemic in 1925/26 , which was associated with 66 deaths , the proportion of non-compliant tanks with Ae . aegypti was highest at 0 . 51 in 1927 ( 1 , 940/3 , 768 , SE ± 0 . 008; Fig 6 ) [43] . From that point , the proportion of non-compliant rainwater tanks containing all mosquito species decreased to 0 . 12 in 1945 ( 284/2 , 344 , SE ± 0 . 007 ) and 0 . 07 in 1952 ( 20/2 , 902 , SE ± 0 . 002; Fig 6 ) , followed by a further rise to 0 . 15 in 1955 ( 359/2 , 443 , SE ± 0 . 007; Fig 6 ) . That increase was probably due to increases in the total number of non-compliant rainwater tanks present during the same period ( Fig 5 ) . As surveys did not identify to species at this time , we cannot determine these were Ae . aegypti or the native container inhabiting species , Ae . notoscriptus . After 1955 , the proportion of infested tanks decreased until the end of entomological surveys in 1971 where container inhabiting mosquitoes were no longer in high prevalence ( Fig 6 ) and non-compliant rainwater tanks were not a common feature in Brisbane’s urban landscape ( Fig 5 ) . To estimate the density of larval habitat provided by domestic rainwater tanks , we calculated the ratio of non-compliant tanks to dwellings in Brisbane from 1912 until 1971 ( Fig 7 ) . High ratios of non-compliant tanks per dwelling ( between 1 . 32 to 0 . 83 during 1912 and 1923 respectively ) relate to periods where epidemic dengue occurred throughout the city . Although the density of non-compliant tanks decreased to 0 . 2 by 1925/26 , the large dengue epidemic during these years may have been facilitated by the high proportion of non-compliant rainwater tanks with Ae . aegypti present ( Fig 6 ) . From 1926 , the ratio of non-compliant tanks to dwellings remained constant , at around 0 . 2 non-compliant tanks per dwelling until 1955 . As Brisbane continued to grow and the number of rainwater tanks started to decrease , the ratio of non-compliant tanks per dwelling decreased to 0 . 08 by 1971 ( Fig 7 ) .
The local extinction of invasive species and the decision on whether an elimination program can be declared successful is challenging . Techniques for detecting invasive mosquitoes are imperfect , and failure to detect a species does not confirm absence with certainty . In Brisbane , despite a constant mosquito management and surveillance program , there have been no recorded specimens of Ae . aegypti outside of first ports of entry or quarantined facilities since 1957 . Surveys conducted by Elizabeth Marks ( Brisbane ) and local health authorities from 1965 until 1980 did not detect the species , and larval surveys in 1995 until 1997 and 2007/08 also failed to isolate the vector in high risk residential areas in Brisbane [69 , 73 , 89–91] . An extensive larval survey of 4 , 983 premises from 2010 until 2012 likewise did not detect Ae . aegypti [92] . Contemporary surveillance efforts have not detected Ae . aegypti in the Brisbane City Council region to date , though the species has been intercepted during routine surveillance at first ports ( Cassie Jansen , Queensland Health , pers . comm 20 May 2016 ) . The absence of locally-acquired dengue cases in Brisbane since 1948 , despite being a notifiable disease , suggests that any population of Ae . aegypti if established , is no longer sufficient in distribution and abundance to vector the virus . The large decrease in dengue fever cases from the 1926 epidemic when compared to the outbreak in 1943 ( Table 1 ) indicates that the mosquito population had likely been reduced substantially prior to the outbreak . As such , the limited cases observed during the 1943 outbreak suggest there was little opportunity for local transmission ( Table 1 ) . Following the large outbreak in Townsville in 1955 ( approximately 15 , 000 cases ) , no cases of locally acquired dengue were recorded in Brisbane , as noted by Doherty [93] who stated that “Brisbane remained unaffected , although each of the seven previous outbreaks of dengue in North Queensland since 1895 had been followed by a high incidence in Brisbane” and from a council annual report , “University collectors have tried in vain to obtain specimens of the vector ( Ae . aegypti ) . This position is a direct result on the active eradication policy of the Council” [94] . These suggest that by 1955 , Ae . aegypti was no longer in numbers large enough to vector the virus amongst the population of Brisbane . Between 1948 and the present day ( despite high numbers of imported cases in recent years [95] ) there have been no locally-acquired cases of dengue fever , suggesting dengue has been eliminated from Brisbane . Our calculations ( based on frequency of sighting records ) suggest that Ae . aegypti is unlikely to be established in Brisbane and it likely went extinct around 1960 . Results from the Jaric and Ebenhard [57] method were not highly significant , suggesting there is a small probability of presence . However our maximum likelihood method , where the probability of detection after the last survey date is the same as before it , suggests it is highly unlikely that the species is present after a period of 59 years . We propose , therefore , that Ae . aegypti is currently either absent from the city or below our ability to detect the mosquito . As Ae . aegypti continues to exist just north of the region ( Fig 1 ) , there is therefore a high risk of future re-establishment . Discontinuing ongoing surveillance could lead to increased future costs in mosquito management and higher disease risk . Historically it is likely Ae . aegypti relied on particular larval habitat features to persist and re-establish through periods of low rainfall [27 , 52] . We believe that three major factors contributed to the disappearance of Ae . aegypti from Brisbane soon after the 1940s . The removal of large domestic water storage containers ( rainwater tanks ) and their replacement with reticulated water has been suggested as the principle reason for the elimination of Ae . aegypti in the city [29 , 45] . O'Gower [27] suggested the species might be eliminated in areas of low rainfall by ‘a continuous mosquito control program , and complete replacement of rainwater tanks by a reticulated water system’ . It is likely that a similar process led to the species disappearance in the Mediterranean region , where commonly-used basement cisterns were gradually replaced with reticulated water [96] . Although Brisbane had reticulated water from the creation of its first dam in 1866 , the crude reservoir systems had no purification and the reservoirs had limited capacity and were prone to the effects of drought . Rainwater tanks were relied upon as a dependable , clean source of drinking water . The first purification system was created in 1919 but tank use continued to increase into the 1950s [79] . It was not until the creation of Somerset Dam in 1954 that a reliable , unrestricted and clean water supply was available to the whole city . This coincided with the period when Ae . aegypti disappeared from the record . The gradual introduction of a clean , reliable source of reticulated water led to the disuse and removal of rainwater tanks throughout the city . Although the removal of these large water storage containers may be considered a tipping point , it was the removal of non-compliant rainwater tanks acting as permanent larval habitat that played an important role in the process of elimination . The specific decline in non-compliant rainwater tanks from the early 1930s appears to coincide with a decline in Ae . aegypti presence and a reduction in dengue transmission . Non-compliant rainwater tanks can provide essential refugia for eggs and larvae and act as population sources after unfavourable climatic periods ( such as droughts , dry seasons and winter ) [27] . It is possible that non-compliant rainwater tanks reflected the general disordered nature of backyards in the early 1920s . The removal of large water storage containers and other larval habitat as a result of house surveys and regulation would have reduced opportunities for population growth and persistence . During peak dengue transmission in the 1920s , ~50% of non-compliant rainwater tanks contained Ae . aegypti . By the early 1930’s , Hamlyn-Harris [97] observed that “never in the history of the mosquito campaign had there been fewer domestic mosquitoes” in the city , although he did not directly record the presence of Ae . aegypti . This is also the period where the ratio of non-compliant tanks to dwellings remained constantly low . This general decrease in non-compliant tanks from 1912 until 1940 may have led to the reduction in container-inhabiting mosquitoes ( Figs 5 and 7 ) , and to the lower incidence of dengue in the 1943 outbreak . During and after the War , a shortage of materials contributed to another rise in non-compliant tanks from 1940 until 1955 which was combated with increased surveillance and enforcement [88] . The deployment of experienced health inspectors after the War [28] and the considerable number of dwellings surveyed from 1948 until 1964 ( Fig 3 ) , was responsible for the subsequent drop in non-compliant tanks . By this time the proportion of non-compliant rainwater tanks containing mosquitoes had decreased substantially when compared to earlier in the century ( Fig 6 ) . By 1952 health authorities understood that Ae . aegypti had been “effectively reduced in numbers to the point where it was no longer a capable link in the chain of [dengue] transmission” [98] . However the elimination campaign continued to “be vigorously waged” [99] in 1964 , and by 1966 the species had been “virtually eliminated” [100] . During 1964 , half the dwellings in the city were surveyed as part of the effort by government officials to eliminate Ae . aegypti . By 1971 , the proportion of non-compliant rainwater tanks compared to total dwellings in Brisbane was minimal . The elimination of Ae . aegypti over large geographic areas is extremely rare . Successful large scale campaigns have focused primarily on top-down approaches , reliant on international interventions and the use of chemical control [33] . The majority of these elimination campaigns were abandoned due to their unsustainable cost , the development of resistance in mosquito populations and the move to less regimented campaigns [55 , 103] . Recently , larval source reduction has been used successfully to eliminate both dengue [104] and Ae . aegypti from villages in Vietnam [36] . This method focused on a very local , bottom up approach that engaged local communities to take ownership of the problem by targeting water storage ( like large cement tanks , wells and ceramic jars ) with a biological control organism , Mesocyclops [36] . This method provided affordable tools for villages to maintain control for long periods of time but unfortunately has not proven applicable to many other environments and has not been widely adopted [36 , 104] . From the inception of anti-mosquito regulations in Brisbane , it was the role of the resident to ensure that their property was not producing mosquitoes . High proportions of breaches in regulations during the period from 1913 until 1955 indicates the initial poor condition of the rainwater tanks in use at that time , and the considerable effort required to reduce the threat they posed . Early in this period ( 1912–1933 ) , surveys for mosquitoes in and around rainwater tanks were performed by trained public health inspectors and entomologists representing state and local government [43 , 71 , 14] . The officials that developed the first mosquito control program were under-resourced and were unable to enforce regulations in the newly amalgamated and growing city [56] . With the formation of the Brisbane City Council mosquito control section , the role of surveillance was undertaken by District Health Inspectors who , by 1942 , were authorized to enforce regulations [56] . The ability to compel householders to rectify breaches in regulations was a necessary response to a very real public health threat . Compliance with anti-mosquito regulations was variable before and during the Second World War , but by 1955 had reached almost 100% . When scaled up to the size of a modern city , a top-down elimination approach is doomed to fail if all stakeholders are not engaged and integrated into the management approach . The public plays a crucial role in mosquito control and elimination campaigns . In 1927 , Hamlyn-Harris stated , “No mosquito campaign which does not include the question of publicity can ever hope to be successful , and this together with the education of the public has become a very important part of daily activities” and “It’s ( Ae . aegypti ) control is only possible provided an educated and sympathetic public cooperate” [43] . We infer that it was the removal of non-compliant rainwater tanks as key larval habitat through the continued enforcement of regulations that played a substantial role in the disappearance of Ae . aegypti from Brisbane . It is likely that alternative mechanisms contributed to the elimination process , however there is little quantitative evidence to show their effect . The elimination of Ae . aegypti from Brisbane incorporated elements of a top-down approach . Unlike the Soper campaigns in South America which used large scale residual chemical spraying to target larval habitat , the strategy in Brisbane was to intensively survey residential areas , and target larval habitat by insisting on screening rainwater tanks and removing all water-holding rubbish [43] . During the decade leading up to 1964 , the majority of dwellings within Brisbane were surveyed for compliance with anti-mosquito regulation and , if in breach , were served with notices to rectify the situation . This top-down intervention identified key mosquito production hotspots and enforced compliance among all members of the population . These measures may be effective in small towns in arid and semi-arid areas where rainfall is limited [27] . However , with the modern proliferation of plastic containers and cryptic subterranean sites ( like underground street drainage ) , the impact we have documented here may remain an artefact of history . It is important that we continue to value the health initiatives that have safeguarded our cities in the past . The methodologies developed in Brisbane over many years show how a successfully integrated regulatory framework , with appropriate public engagement and enforcement can be applied to mosquito control . These historical records of management processes are rare , and this example shows how Ae . aegypti may be dramatically impacted at the margins of its range if key larval habitats are managed appropriately and enforced by effective legislation . Drought conditions during the early 2000s have resulted in the installation of over 300 , 000 rainwater tanks in Queensland ( >41% of properties in Brisbane ) . This has been driven by a culture of water harvesting via educational campaigns and legislated water restrictions [117] . Likewise , large numbers of water barrels and cisterns have been installed as a human response to drought throughout southern California and are likely to be an ongoing concern for local health authorities monitoring the spread invasive mosquito species [118] . As such , these regions should acknowledge the potential role of these behaviours on the reintroduction and spread of Ae . aegypti . A clearer understanding of the threat posed by past epidemics and the enormous effort required to eliminate such a threat provides justification for ongoing surveillance by local and state authorities to ensure Australian cities remain vector free . With the current distribution limit of Ae . aegypti just north of Brisbane , ongoing mosquito surveillance and rainwater tank monitoring is essential . Although regulations are still in place , little is being done to inspect the condition of rainwater tanks , enforce anti-mosquito regulations , or educate residents on proper tank management within the Brisbane council area . The emergence and re-emergence of arboviruses including dengue , chikungunya and Zika viruses across the globe , further highlights the importance of early detection and response to invasive urban vectors . The successful campaign that led to the elimination of Ae . aegypti from Brisbane offers insights into the challenges we may face in the future , and provides an important source of knowledge from which to plan for and combat invasions of disease vectors internationally . | We examined the historical role that water storage practices and the enforcement of anti-mosquito regulations played in the elimination of Aedes aegypti from Brisbane , a major urban centre in Australia . We examined changes in regulations pertaining to mosquitoes , collected government records documenting surveillance , and the response by the community to the actions of local authorities . Our findings indicate that anti-mosquito regulations , underpinned by effective implementation , were successful in gaining community support and removing the risk of mosquito presence at non-compliant properties . In particular , we argue that the removal of rainwater tanks which provided a permanent larval habitat in otherwise suboptimal environments , played a major role in the elimination of the species from Brisbane . Public Health regulations were supported by a large surveillance effort by local government health officers that were empowered to enforce legislation where necessary . Our findings are of importance to health authorities managing the ongoing expansion of Aedes populations , particularly in regions of sub-optimal climate and where water storage has become a major concern . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"invertebrates",
"medicine",
"and",
"health",
"sciences",
"ecology",
"and",
"environmental",
"sciences",
"water",
"resources",
"geographical",
"locations",
"social",
"sciences",
"australia",
"animals",
"ddt",
"insect",
"vectors",
"zoology",
"law",
"and",
"legal",
"sciences",
"conservation",
"biology",
"infectious",
"diseases",
"agrochemicals",
"regulations",
"natural",
"resources",
"aedes",
"aegypti",
"disease",
"vectors",
"insects",
"agriculture",
"arthropoda",
"insecticides",
"people",
"and",
"places",
"conservation",
"science",
"mosquitoes",
"species",
"extinction",
"entomology",
"oceania",
"biology",
"and",
"life",
"sciences",
"species",
"interactions",
"evolutionary",
"biology",
"evolutionary",
"processes",
"organisms"
] | 2017 | The elimination of the dengue vector, Aedes aegypti, from Brisbane, Australia: The role of surveillance, larval habitat removal and policy |
Classification of plants according to their echoes is an elementary component of bat behavior that plays an important role in spatial orientation and food acquisition . Vegetation echoes are , however , highly complex stochastic signals: from an acoustical point of view , a plant can be thought of as a three-dimensional array of leaves reflecting the emitted bat call . The received echo is therefore a superposition of many reflections . In this work we suggest that the classification of these echoes might not be such a troublesome routine for bats as formerly thought . We present a rather simple approach to classifying signals from a large database of plant echoes that were created by ensonifying plants with a frequency-modulated bat-like ultrasonic pulse . Our algorithm uses the spectrogram of a single echo from which it only uses features that are undoubtedly accessible to bats . We used a standard machine learning algorithm ( SVM ) to automatically extract suitable linear combinations of time and frequency cues from the spectrograms such that classification with high accuracy is enabled . This demonstrates that ultrasonic echoes are highly informative about the species membership of an ensonified plant , and that this information can be extracted with rather simple , biologically plausible analysis . Thus , our findings provide a new explanatory basis for the poorly understood observed abilities of bats in classifying vegetation and other complex objects .
When orienting in space and searching for food , microchiropteran bats continuously emit echolocation signals . The returning echoes are analyzed in the auditory system to perform the basic echolocation tasks of detection , localization and classification [1] . Classification of vegetation probably plays a major role in spatial orientation and in food acquisition . It is fundamental for recognizing landmarks and vegetation edges which are mandatory for the route following behavior observed in bats [2] . In addition it is also very important for finding and recognizing foraging habitats such as meadows , bushes , trees etc . which are indicators of specific food sources [3] , [4] . In all of these cases the vegetation has to be classified from a relative long distance of up to a few meters . The behavior of bats in the field indicates that bats notice background structures within the so called edge space which extends up to around 6 m [5] . It has also been shown that Natterer's bats learn to discriminate conifers from broad-leaved trees and that horseshoe bats commuting along a hedge of bushes show distinct reactions in their echolocation behavior when the reflection properties of the bushes are changed by covering them with velvet ( Denziger and Schnitzler , unpublished data ) . In addition to the classification of vegetation types , bats can also identify parts of plants like flowers and fruits . Glossophagine bats for instance , find new nectar sources by classifying the shape and texture of flower echoes [6] , [7] . Plants have complex shapes that cannot be described in terms of simple geometrical primitives [8] . From an acoustical point of view , a plant can be approximated as a stochastic array of reflectors formed by its leaves . McKerrow et al . [9] removed the leaves from pot plants and discovered that the contribution of branches to the echoes is minor . In large plants the stem might also play a role . In broad-leaved plants , the reflectors are relatively flat and usually large compared to the emitted wavelengths ( ∼0 . 3–1 . 5 cm ) in a typical frequency modulated bat call . Hence , the backscatter from a broad-leaved plant typically is a superposition of reflections , with statistics determined by the characteristics of the foliage such as the size and the orientation of the leaves , along with their spatial distribution . The overall duration of the echoes is a result of these parameters too . In dense foliage , for instance , surface leaves will acoustically shadow deeper ones , thus strongly attenuating the sound waves that penetrate beyond the outer surface . These properties also apply to conifer trees , except for the fact that they possess needle-shaped reflectors that are small relative to a considerable part of the emitted wavelengths . Conifers are therefore regarded as diffuse scatterers that produce many small echo components , whereas broad-leaved plants lead to pronounced amplitude peaks in the echoes , referred to as glints . Although the importance of classifying complex objects is well discussed in the scientific bat literature , very little is known about how bats actually perform classification . Only a few previous studies directly addressed the question of object classification using echolocation in bats , and most of them did so in the context of classifying objects with rather simple shapes [10] , [11] , [12] , or only a few reflectors [13] , [14] . The few experiments that tested the bat's ability to classify relatively complex echoes [15] , [16] did not suggest an explicit mechanism to explain it . The studies that examined classification of simple objects usually assumed simple cues that could be easily recognized in the temporal , frequency or time-frequency representation of the echoes as a basis for classification such as , for instance , a certain notch arrangement in the frequency domain . This approach is hardly feasible for real plant echoes due to their complexity and the strong dependency on the angle of acquisition which makes the ad hoc identification of such features a difficult task . Another typical approach is to identify peaks corresponding to reflections from parts of the object and to compare them to stored echoes that represent known objects or known geometrical shapes ( e . g . , edges , corners and surfaces ) . The comparison can be done by measuring the difference between the echoes directly [15] or by comparing certain representative statistics [17] . Once again , these methods will face severe difficulties with complex echoes , mainly since the echoes returning from different reflectors always highly overlap and are very hard to isolate . A few studies trying to classify complex echoes such as vegetation echoes [13] , [18] and Stilz and Schnitzler unpublished data relied on extracting one or several parameters ( e . g . peak intensity , average intensity and etc . ) from some representation of the echoes , with a subsequent selection of those parameters that best assign the plant echoes to their corresponding classes . Thus , the set of all tested parameters is determined by the experimenter beforehand . This has advantages and disadvantages: on the one hand , parameters are usually chosen according to physical or biological plausibility which simplifies their interpretation , but on the other hand strong assumptions are made by choosing a fixed set of candidate parameters since some of the important features might be overlooked . In this paper , we propose a new approach to complex echo classification . We use a linear classification technique that comes originally from the field of machine learning . We use this technique to operate directly on the raw spectrogram magnitude of the echoes , without the intermediate step of specifying some set of potentially relevant parameters or features . With this approach we take advantage of the statistical structure of the data itself in order to identify the best features to classify it . Thus , the technique allows for the exploration of a wide range of features simultaneously , and often finds simple ones . This comes at the price that the obtained results are slightly harder to interpret on first sight , but we will provide a thorough analysis of the features that are extracted from the data . Our classifiers are trained on a large database of natural plant echoes , created with a bat-like ultrasonic frequency modulated signal . We show that the trained classifiers are able to classify echoes from previously unseen plants with high accuracy . At the same time , our method provides a systematic analysis of all linear features in the echo spectrograms of the database in terms of their relevance for classifying the underlying plant species . More over our approach enables classification of vegetation echoes using a single echo . This coincides with recent work [14] that showed that bats can classify a complex 3D object using a single ensonifying position , without the need to integrate the information from echoes over different acquisition angles . The presented approach provides many insights regarding the task of plant echo classification and is sufficiently general to be applied to other types of complex echoes , for instance from food sources or landmarks .
A linear SVM classifier is able to distinguish between any of the five tested plant species and any other species or group of species , based on a comparison between two single echoes , one from each class . For the classification task of discriminating one species from the rest already a simple linear classifier achieves very high percentage of discrimination ( 80–97% , see Table 1 for details ) . The classification of spruce or corn from the other species is almost perfect whereas the classification of the three broad-leaved trees , and especially the beech , from the rest was the most difficult . For the pairwise classification ( Table 2 ) the relatively poor result for the classification of beech vs . blackthorn , both broad-leaved trees , stands out . The relatively high standard deviation in this case implies that a larger data set might improve performance . Comparing the task of pairwise classification in general to the task of one species vs . the rest reveals that the latter is the more difficult one . This is expected since a group of species always contains much more intrinsic variation that the classifier has to learn , but even with this difficulty , our linear classifiers performed surprisingly well . In the next sections we will mainly discuss the task of classifying one species against the rest , except for cases in which the pairwise comparison reveals more interesting phenomena . The weights of the normal vector to the separating hyperplane , i . e . , the decision echo , has the same dimensionality as the data , and can assist in better understanding the features that are used by our machines for classification . Since we are using linear machines , the class of an echo is actually determined by the sign of the inner product of the preprocessed echo and the decision echo , after adding the offset . This means that the regions of the decision echo that have high absolute ( depicted dark or bright in the figures ) values have more influence on the decision . In order to interpret the decision echo , we present the decision echoes of the classification tasks of spruce vs . the rest and corn vs . the rest together aside an image of the difference between the average spectrograms of the two classes ( Figures 1 and 2 ) . Comparing the decision echoes and the spectrogram differences ( Figures 1C and 1D , 2C and 2D ) it becomes clear that in both classification tasks our classifiers are actually emphasizing the areas in which the differences between the spectrograms are most salient . The comparison of the differences between the decision echoes of the two tasks shows that in the task of classifying spruce from the rest , the classifier performs a combination of a frequency domain analysis and a time domain analysis . In the early parts of this task's decision echo , low frequencies are inhibitory ( with negative values ) while the high frequencies are excitatory ( with positive values ) . In the later parts ( ∼ after 10 ms ) the entire decision echo is excitatory ( excluding regions with larger attenuation as will be explained below ) . Therefore , classification of spruce can be generally described as a measurement of the difference between the high and low frequencies intensities in the spectrogram's early parts ( frequency domain analysis ) and as a measurement of all intensities in the later parts ( time domain analysis ) . The classification of the corn field is mainly a time domain analysis . Here the regions in the decision echo which are compatible with the first and second rows of the field ( compare with the corn spectrogram in Figure 2A ) are excitatory , while the gaps between these rows are inhibitory . The effect of the frequency dependent atmospheric attenuation of sound waves is expressed in all of the decision echoes . According to this attenuation , the higher the frequency of the wave is , the faster its intensity decreases with the distance . This gives the decision echoes a triangular shape , meaning that the higher the frequency , the less the later parts of the spectrograms are used for classification ( gray regions in Figures 1 and 2 ) . An alternative interpretation of the decision echo is the direction in the high-dimensional input space along which the changes between the two classes are maximal . In other words , for a pair of species it represents the transition between the two . Inspired by Macke et al . we calculated for each pair of species the average spectrogram , and then added the decision echo multiplied by a positive or negative factor η . By doing this we actually move along the direction of the maximum change from a mean representation of the two plants in the directions of each one of them . We used this method to generate 1000 artificial spectrograms that are hybrids of different ratios of the apple vs . corn pair ( 500 on each side of the hyperplane see Figure 3 ) . To generate echoes from the hybrid spectrogram , we propose to use the random phase method described in the Materials and Methods section . We did so in order to verify our method , and the resulting echoes lead to a consistent classification behavior , i . e . , higher classification performance for larger absolute values of η ( see Figure 3 for more details ) To determine the separating hyperplane , the SVM uses only a limited number of data points ( the ones that are closest to the hyperplane ) which are termed support vectors . The importance of the ith support vector is weighted by a constant αi . Adding up the support vectors on each side of the hyperplane separately , with the proper weighting , provides another view on the classification rule . For an arbitrary pair of two species , a weighted sum of the support vectors on one side of the hyperplane can be intuitively understood as the most similar this species can acoustically be to its pair in the limits of our data set . The spectrograms of the weighted support vectors for the pair of apple tree and corn field reveals how in some cases an apple tree can acoustically resemble a corn field and vise versa ( Figure 4 ) . From the decision echoes we learned that both time and frequency information are used for classification and that in higher frequencies the earlier parts of the spectrograms are preferred for classification , probably due to atmospheric attenuation . Here we test whether classification is possible when only parts of the spectrogram's information are used . We divided the spectrograms into squares of 5 kHz by 5 ms , and for each square , we trained and tested SVMs for all the classification tasks in the same manner described above . We found that already the information contained in one of the limited squares within the spectrogram is sufficient for classification with very high ( ∼0 . 9 ) performance in all cases except for beech ( Figure 5 ) . However , the exact position of this limited sensitive region in the time-frequency space can be significantly different for different classification tasks . In spruce classification for instance the low frequencies in the beginning of the echo provide the best classification performance . In blackthorn on the other hand the later parts of the spectrogram are better for classification , and there is a wide range of frequencies and times that can be used with almost equal performance . Our classifiers generalized over different aspect angles . This can already be learnt from the basic experiments since we trained them by using data from all angles , and then tested them with high success on data from all angles ( Tables 1 , 2 ) . In a different version of the one species vs . the rest experiment we trained machines using training data recorded from all angles except for the tested one and then tested on data points from only the tested angle . The classification performance in these experiments stayed as high as in the ones in which data from all angles were used to train and test the machines with no significant difference ( Two way ANOVA , F2 , 60>0 . 86 , P<0 . 45 ) . In order to examine the sensitivity of the performance of our machines to the preprocessing of the data , we used a cross-validation approach to estimate the performance while changing the parameters of the preprocessing steps . This was done on the training data set as explained in the methods section for two procedures: the effect of cutting out the echoes in the time domain , and the effect of the time-frequency resolution ( i . e . , the DFT window length used to calculate the spectrogram ) . To test the effect of cutting the echo out in the time domain , we changed the threshold according to which the cutting points were determined . Cutting the echo improved the classification performance by a non significant average of 0 . 02 ( Two way ANOVA , F2 , 60>1 . 78 , P<0 . 18 ) We attribute this slight improvement to the registering effect that this procedure has on the echoes . Applying a threshold is closely equivalent to recognizing the first wave front of the echoes and this aligns them before any further processing . The two different cutting criteria ( 10 or 20 times above noise level ) showed no difference what so ever . To determine the effect of the DFT window length we varied it and kept the percentage of the overlap between sequential windows constant ( Figure 6 ) . The extent of the spectrograms in the temporal direction decreased with window length whereas the extent in frequency increased such that the overall information remained constant . Up to a certain window length ( 1000 ) , representing a time bin of 1ms ( with 80% overlap ) the window length had no significant influence on classification performance . Above this length however , for the 2000 window , there was an overall significant decrease ( 0 . 07 on average ) in classification performance ( 2-way ANOVA , F3 , 80>18 . 5 , P<0 . 0001 ) . This decrease mainly affected the three classification tasks blackthorn vs . rest ( 0 . 25 on average , 1-way ANOVA , F3 , 16>24 . 8 , P<3−6 ) , beech vs . rest ( 0 . 13 on average , 1-way ANOVA , F3 , 16>6 . 5 , P<0 . 005 ) and corn vs . rest ( 0 . 03 by average , 1-way ANOVA , F3 , 16>2 . 85 , P<0 . 07 ) while the performance of the other two tasks did not change . The decrease is probably a result of the loss of time information due to excessive smoothing . In general , the most suitable window length depends on the specific classification task .
In this work we analyzed the characteristics of a database containing vegetation backscatter from five plant species ensonified with a bat-like ultrasonic pulse from different aspect angles . We used a linear classification technique to find discriminative features in the backscatter spectrograms that were able to differentiate between different plant species independent of aspect angle . In contrast to previous approaches , we did not derive these features from biological or practical plausibility assumptions . Instead , discriminative features were learned from the statistical regularities found in our database . When we tested our classifiers on a single echo from a new , previously unseen specimen from one of the species in the database , classification performance was surprisingly high , ranging between 0 . 8–0 . 99 . This indicates that the echoes created by a frequency modulated ultrasonic sweep can be highly informative about the plant's species membership . This forms a possible explanatory basis for some of the observed abilities of bats in classifying complex objects such as landmarks or vegetation as indicator for food sources [3] , [4] . Once a linear classifier is trained , it can also be used as a generative model . This means that the learnt features can be used to generate new artificial examples of the data . In our case we could create new echoes of a certain plant species or of a combination of species ( Figure 3 ) . In the future we hope to use this type of artificially generated echoes in behavioral experiments in order to test the correlation between our linear functions and bat classification performance . As described in the methods , we designed our preprocessing procedure in such a way as to minimize the species-specific noise ( due to external or internal recording parameters ) to prevent the classifiers from using it for classification . The probability that such artifacts still retain some influence on our results is quite low considering the actual information that leads to a classification decision as depicted in the decision echoes . All decision echoes ( see examples in Figures 6 and 7 ) give a higher weight to regions of the spectrogram where the signal of at least one of the classes is high above the noise level . Regions with lower signal intensities , i . e . later in time and higher in frequency , tend to have values close to zero in the decision echoes . As an additional test , we repeated the same classification experiments , but this time after preprocessing the echoes with a Wiener filter [19] , which uses the noise spectrum in order to filter out the noise from the entire signal , not only from the low amplitude regions . The noise spectrum for each echo was estimated in the same way as described in the methods . There was no significant difference in the classification performance of the classifiers with and without Wiener denoising ( F1 , 48>1 . 6 , P<0 . 22 ) . The results after denoising appear to be slightly ( but not significantly ) better which implies that the measurement noise does not contain species-specific artifacts that could be erroneously used by the algorithm for classification . When examining the decision echoes it seems that some of them ( e . g . corn classifiers , see Figure 2 ) use the time structure of the echoes more than the frequency content , while others ( e . g . spruce classifiers , see Figure 1 ) use the frequency content more than the time structure . In general , in all cases both time and frequency information was used for classification . Regarding the best features of the plants used for classification , it seems that our classifiers neither use the overall extent , nor the fine texture of the spectrogram . Instead they rely on intermediate scale structures , such as the representative frequency content in a certain time interval or a characteristic time structure for certain frequencies . In most cases we could identify a small region in the spectrogram which is already sufficient for classification . However , the exact position of this decisive region in the time-frequency plane can significantly change between the different classification tasks . This means that if nothing is known about the classified plant species beforehand , a large proportion of the spectrogram is required to achieve a good performance over all tasks . Thus , a call with a large frequency bandwidth , as is observed in frequency modulating bats , is preferable from the classification point of view . A plant is a complex object comprised of many reflectors ( mainly the leaves ) . Although the spatial arrangement of the different plant species contributes to the echo structure , it can be helpful to regard the plant leaves as an array of independent , rather simple reflectors to understand the differences in the frequency content of species . In our study we found that the most suitable frequencies for classification are not necessarily the ones with the best signal to noise ratio ( SNR ) . The highest SNR was usually attained around 50 kHz , whereas the frequencies with the best classification performance were in most cases lower , indicating that the echoes vary more in the lower frequency range between species . Some reason for these preferred frequency bands can be found in radar theory [20] . The cross section of a reflector depends on the geometry of the reflector in relation to the wavelength of the sound pulse . For a simple spherical reflector , the intensity of the echo depends on the ratio between the sphere's circumference and the wavelength of the emitted signal . This ratio defines three regions: ( 1 ) The Rayleigh region - if the circumference is smaller than the wavelength the intensity of the reflections decreases rapidly when decreasing the radius of the sphere . ( 2 ) The resonance region - if the wavelength is of the same order as the circumference ( up to ∼10 times larger ) the intensity of the reflection oscillates depending on the ratio mentioned above . ( 3 ) The optic region - if the circumference is much larger than the wavelength the intensity of the reflection is equal in all frequencies . This division into three domains exists also in reflectors with a more complex shape , but then the cross section will also depend on the angle of ensonification . The borders of these regions when considering the extreme frequencies of our emitted signal ( 25 and 120 kHz ) are such that reflectors larger than 14 cm will be in the optic region for all frequencies , and reflectors smaller than 0 . 03 cm will be in the Rayleigh region for all frequencies . The reflectors in between will be in all three regions depending on the frequency . From the point of view of classification , it is clear that the Rayleigh region is the most advantageous since at a given frequency , the intensity of the reflection changes with the circumference , therefore providing direct information about the reflectors size . Clearly , this presupposes that the intensity is high enough to be perceived . The optic region on the other extreme provides no frequency information that could be used for classification , since the reflections in all frequencies are redundant . Obviously , the time structure can still be different . The resonance region shows a more complex interdependence between frequency and reflector size than both extremes , but a suitable classifier might be able to use this information . In order to relate this theoretical framework to our data , we have to provide some approximation of our reflector's circumference . This is not easy , for the leaves on plants comprise of a range of many sizes , and they are not simple spheres . In the case of spruce , its needles prevent us from doing this , but it is safe to assume that it's very small radial dimension ( up to a few millimeters ) is equivalent to relatively high frequencies , above 100 kHz , and therefore most of its reflectors will behave according to the Rayleigh domain . Corn leaves on the other extreme are very long , and will therefore probably mainly behave according to the optic domain . As for the three broad-leaved trees , we use the roughly approximated average leaf length ( calculated by measuring a variety of leaves ) in order to estimate the relevant wavelength range . Apple and beech trees exhibit the largest leaves among the three , with an average length of around 8 cm . This is equivalent to a wavelength of a few kHz . Its reflectors should therefore behave according to the resonance domain when the emitted signals have frequencies of up to a few dozens of kHz , and according to the optic domain with higher frequencies . Blackthorn trees exhibit smaller leaves , with an average length of about 3 cm . This is equivalent to a wavelength of roughly 10 kHz , resulting in its reflectors being in the resonance domain for most of the frequencies of the signals emitted in this research . Spruce classification is probably easiest to explain by to this approach . Its many reflectors in the Rayleigh region result in lower intensities in the low frequencies of its echoes ( Figure 2 ) . This means that it can be well classified by its lack of low frequency content . Indeed , as can be seen in the decision echo and time-frequency classification performance ( Figures 3C and 7A ) , the information in low frequencies provide the best classification performance for spruce . Corn field in contrast should not contain much frequency information , and truly its decision echo doesn't seem to be using any obvious frequency information ( Figure 2D ) , and so does the time-frequency classification performance graph imply ( Figure 5D ) . In the case of the three broad-leaved trees ( apple , beech and blackthorn ) the effects of frequency are less obvious . We therefore examined the classification performance of each pair when only using parts of the spectrograms with a limited bandwidth of 10 kHz while retaining the entire time information . For all pairs , classification was best at low frequencies ( Figure 7 ) . For beech vs . blackthorn and apple vs . blackthorn , all frequency bands between 25–80 kHz lead to a similar classification performance , whereas in beech vs . apple , performance begins to drop already at the 30–40 kHz band . These could be explained by the above argumentation: all three plants exhibit leave sizes in a considerable large range such that for our emitted call all three species probably have reflectors both in the resonance and in the optic regions . Apple and beech trees , however , have bigger leaves than blackthorn and thus should have more reflectors in the optic region and less in the resonance region , particularly at higher frequencies . As a consequence , apple and beech should be harder to discriminate in this frequency range . Since the intent of our study is to test which features of plants echoes might enable bats to classify the plants , we have to examine if the information used by our classifiers is – at least in principle – available to the bat brain . After the preprocessing of the received echoes our classifiers were trained to recognize plant species based on the magnitude of their spectrograms . This information is easily accessible to the bats through the spectro-temporal decomposition of the echo in the cochlea [21] . We ignored the phase information which to date has not unequivocally been proven to be used by bats . We also did not cross-correlate the recorded echoes with the emitted signal . This is often done in echolocation studies , thus revealing the impulse response ( IR ) of the ensonified object , although it is not known whether bats can actually use the IR . Finally , we use a time resolution of about 1ms which is far above the minimum time resolution which has been reported for bats [22] , [23] . Thus it seems highly probable that the information used by our classifier is available to bats . Experimental evidence suggests that bats can extract information with a much higher resolution than required ( see [23] for a summary ) . The classifiers were able to classify a plant correctly at acquisition angles that were not present in the training set , i . e . , our classifiers generalize to a certain degree over the angle of acquisition . This result was unexpected , since in acoustics , as opposed to vision , a slight change of the acquisition angle can result in a very large change in the echo , as has been shown for plants [9] , [11] . However , we noted above that our classifiers use intermediate-scale features which probably vary more slowly over the angle of acquisition . Moreover , most of the species in our database contain leaves in all orientations such that the local statistics do not change significantly with acquisition angle , even when the individual echoes vary considerably . An issue that was not tested in this work is the generalization over distance , i . e . the ability to use the same classifiers on objects that were ensonified from different distances . The two main limiting factors regarding this generalization are the attenuation of the echoes and the change of the beam width . The attenuation affects the echoes in two ways: 1 ) The SNR of the entire echo deteriorates , in a frequency dependent manner . 2 ) The geometric attenuation increases with the square of the distance , and therefore the attenuation rate within the echo will change when it returns from different distances . The first problem of the overall SNR could be dealt with , up to a limit , by increasing the intensity of the emitted signal . In addition , our classifiers do not require the fine texture of the spectrograms for classification , and therefore can probably tolerate a certain deterioration of the SNR without a significant drop in performance . The second problem could be overcome – at least in principle – by using the absolute distance as measured by the arrival time of the echo to compensate for the attenuation differences within the echo . As for the beam , its width will widen the further the emitter is from the plant , thus increasing the ensonified region . The larger the emitter distance , the more reflectors will contribute to the echoes . Taking into account the intermediate features used by our classifiers , we hypothesize that as long as our beam is wide enough to capture them , classification performance will stay high . A too wide beam , however , could introduce new echoes from other reflectors , which leads to a smearing effect due to the arrival of more reflections at close instants in time , and thus to a slow deterioration of classification performance . Although bat beams are usually much wider than the one used by us , it is clear that there exists a distance range in which the echo statistics are similar to our setting . In one of the few reported works dealing with the bat's ability to classify complex echoes , Grunwald et al . [14] found that bats can distinguish the fourth moment of artificially created echoes . They conclude that bats might be using the changes in the fourth moment to facilitate navigation guided by echolocation . We tested this conclusion in the light of our results for two pair-wise classification tasks . To this end we calculated the fourth moment of each echo and compared it to its distance from the hyperplane ( see methods ) . The results ( Figure 8 ) show that in the rather simple task of classifying a conifer tree ( spruce ) from a broad-leaved tree ( apple ) the distance from the hyperplane of each echo is linearly correlated with its fourth moment ( R∼ = 0 . 64 , P<0 . 00001 ) . However , since we were using only linear machines , our classifiers have no access to higher order statistics such as the fourth moment . This means that information sufficient to classify the two trees is also available in the low order statistics of the echoes . In the case of a difficult classification task ( blackthorn vs . beech ) on the other hand , we found a close to zero linear correlation between the distance from the hyperplane of the echo and its fourth moment ( R∼ = 0 . 1 , P<0 . 00001 ) . Moreover when examining the data ( Figure 8B ) it is obvious that only the fourth moment is not a sufficient statistic for discriminating between these two broad-leaved tree species . In contrast , the SVM is able to find features that are sufficient for reliable classification of this pair already by relying on simple first- and second-order statistics . Wichmann et al . have shown the relevance of a hyperplane calculated from the data to human categorization performance [24] , [25] . They compared SVM-based classification with human performance on a task of image gender classification , and found that SVMs are able to capture some of the essential characteristics used by humans for classification . Furthermore , Wichmann and Macke were able to show that the distance from the separating hyperplane could be used to predict the certainty with which these decisions are made . Despite the fact that it is known that the brain can perform classification of nonlinear data , these works always used linear machines just as we did . In the future we would like to use the SVM as echo generators in order to test the relevance of our calculated hyperplanes to performance of the bat brain . We have found that the highly complex echoes created by ensonifying plants with a frequency modulated bat like signal contain vast species specific information that is sufficient for their classification with high accuracy . From the point of view of a bat , we prove that it can use a single echo received by one ear , with a surprisingly simple receiver , having a relatively low time resolution and no access to the impulse response , to extract the information required for classification . We also demonstrate how it can then apply a basic linear hyperplane that could be easily implemented by a neuronal apparatus , in order to classify the vegetation echoes . These findings could explain some of the abilities observed in natural bat behavior such as using landmarks for navigation , and finding food sources on specific vegetation .
A biomimetic sonar system consisting of a sonar head with three transducers ( Polaroid 600 Series; 4-cm-diam circular aperture ) connected to a computer system was used to create and record vegetation echoes . The sonar head was mounted on a portable tripod . Its central transducer served as an emitter ( simulating the bat's mouth ) and the two side transducers functioned as receivers ( simulating the ears ) . Backscatter received from the emitted signal was amplified , A/D converted , and recorded by a computer . The emitted signal resembles a typical frequency modulated bat call in terms of its duration and frequency content ( Figure 9A ) . It comprises a four millisecond linear down-sweep from 140 to 25 kHz . We excited the emitter with a constant amplitude , but due to the speakers frequency response an uni-modal response function was created with a maximum around 50 kHz , providing an intensity of 112 dB ( SPL ) at the maximal frequency in a distance of 1m from the emitter . Most of the signal energy was contained in the frequency band between 25–120 kHz . The combined frequency response of our emitter and receivers resulted in a frequency response that resembles the one of a typical frequency modulated bat call . In contrast to bats our emitted sound pulse had a rather narrow beam width , with its first null for 50 kHz occurring around 15° , much lower than known for bat calls [26] . The recorded back scatter or echo ( both terms will be equally used in this paper , Figure 9B ) was digitized at a sampling rate of 1 MHz and with a 12-bit resolution . The length of the recorded echo was very long ( 40 ms corresponding to 6 . 8 m ) . It included a long tail of noise after the part with echoes returning from the target . This enabled exact estimation of the noise for each recording . All recordings were performed in the field with real plants as targets . Five plant species were chosen , representing a variety of the common species available in the local bats environment . The species were: 50 specimens of each species were ensonified , each one from 25 different aspect angles on an equally spaced 5×5 grid centered at the horizon and the midline of the tree . This was done by starting at the top most left point on the grid , 10 degrees above the horizon and 10 degrees left to its midline and then turning the sonar head right in sequential steps of 5 degrees along the 5 points of the first row . Next the head was lowered by 5 degrees and the procedure was repeated , this time towards the right . This procedure provided 1250 echoes for each species from each ear . The distance between plant and tripod was always 1 . 5 m , and the height of the tripod above ground was set to 1 . 35 m . The acquisition of data from different angles enabled us to test for the ability to identify species independent of the aspect angle . This is commonly done in image classification research [27] , in which images of the same object are taken from different angles in order to test view point invariance . All of the signal processing was performed with matlab 7 . 0 The recorded echoes went through several three preprocessing steps . For all training experiments described in the following paragraphs , the data was divided into a training ( four fifth ) and a test set ( one fifth ) . This was done such that all the angular echoes of a specific plant individual were attributed either to the test or to the training set , but never to both , to prevent leakage of information from the test set to the training set , which might result in an overestimation of the generalization performance . We performed two kinds of classification experiments . The first was a pairwise classification in which we trained ten machines , to distinguish between any possible pair of species . In the second , we trained five machines , each capable of distinguishing between one species and the other four . It should be mentioned that our classifiers categorize the plant using only a single echo . This is different from all the previous plant echo classification studies . After applying the above preprocessing methods , with a DFT window of 1000 , each echo was represented by a 95 ( frequency bins ) ×90 ( time bins ) = 8550-dimensional spectrogram , assuming here that the 1000 point window was used . Next each spectrogram was rearranged as a 8550-dimensional vector ( simply by concatenating its columns ) which left us with a total of 6250 echoes , each represented by a 8550-dimensional vector . We used Principle Component Analysis ( PCA ) to reduce the dimensionality of the data before applying the machine learning algorithms . We did this by projecting each data vector on the 250 eigenvectors with the highest eigenvalues . In every experiment , the eigenvectors were calculated for the covariance matrix of the training set exclusively . As a common PCA pre-process all 8550-dimensinal data vectors were first normalized to have equal energy . The PCA transformation reduced the dimensionality of the data so that each echo could now be represented by a 250-dimensional vector . The number 250 was another free parameter that was chosen via cross-validation ( see below ) . We used linear Support Vector Machines ( SVM , [29] , [30] , Figure 9D ) as our classification algorithm . To implement the SVM we used the free “spider” software ( http://www . kyb . mpg . de/bs/people/spider ) . An SVM is a state-of-the-art learning algorithm based on statistical learning theory . A linear SVM can be intuitively interpreted in a geometrical way as a separating hyperplane that divides the data set into two classes by minimizing the classification error of a training set and at the same time by maximizing its distance from the data points that are closest to it ( Figure 9D ) . The hyperplane is simply a multidimensional plane that has the same dimensionality as the data points which correspond , in our case , to the spectrograms of the echoes after the above preprocessing . In many cases a perfect separation of the data into two classes is not possible due to outliers , or due to an overlap of the classes . Therefore the learning algorithm is adjusted to enable a certain amount of misclassified points . For this purpose a new constant C is introduced , that defines the penalty for misclassified points . This constant is known as the free parameter of the SVM - and as the other free parameters - it was determined by cross validation . After training the SVM , classification was performed according to the following calculation:where is a vector normal to the hyperplane is a test echo ( after preprocessing ) and b is an offset ( also calculated by the learning algorithm ) . The offset is equivalent to the free parameter in a three dimensional plane , and changing it moves the hyperplane along its normal direction When the result is +1 the echo will be classified as belonging to one species or species group and when it is −1 it will be classified as belonging to the other species or species group . It should be noted that SVMs is a non parametric method that makes no prior assumptions on the data and learns the classification rule using the data itself . The normal vector of the hyperplane is a weighted linear combination of the training data points: ( 2 ) Where yi is the sign ( ±1 ) attributed to each training data point according to its class label . The weights αi are a result of the learning procedure , and for most points they will be zero . Only the points that are closest to the hyperplane on both sides are assigned non-zero weights . These points are called support vectors , and actually define the orientation of the hyperplane . They can be interpreted as the most difficult points to separate in the limits of the data set . In visual classification studies the normal vector is interpreted as the decision-image [24] , [25] , so we will call it in our context the decision-echo . The decision-echo can assist in better understanding the features that are used by the machine for classification . It has the same dimensionality of the data points after preprocessing , and since we were only using linear machines , the class of an echo is actually determined according to the sign of the inner product of the echo and the decision echo added to the offset . This means that the regions of the decision echo that have high absolute ( non zero ) values are more important for classification . An alternative interpretation for this vector is the direction along which the change between the two classes is maximal . In addition to classification , one can calculate for each echo its distance from the hyperplane by: ( 3 ) This measurement provides additional information regarding the ordering of our data points according to the classifier and can be used for further understanding of our performance . The four parameters of our model ( i . e . , the threshold above noise for the cut in time domain , the DFT window length , the number of principal components for projection and the C parameter of the SVM ) were all determined using a five-fold cross validation . This means that for each possible value of the parameters , the training data set is divided into five sets of equal size , and each set serves as a test set for a classifier trained with this specific value on the other four sets . The value yielding the highest average classification rate was then chosen ( see performance measurement below ) . It is important to note that this procedure was executed exclusively on the training set . The first parameter – the threshold above noise level ( step 1 of preprocessing ) was determined independently of the other three , after they were already set . For this parameter the values 1 , 10 and 20 times above noise level were tested . The latter three parameters were determined via a cross validation on a 3-dimensional grid of parameter combinations . This means that for each possible combination of the free parameters on the grid the cross validation procedure was executed . The combination yielding the highest average classification rate was then selected . The possible values for these three parameters were as following: In the case of the window length the values 250 , 500 , 1000 and 2000 were tested . For the dimensionality reduction via PCA we tested the values 150 , 200 , 250 and 300 principle components and the values for C were evenly chosen on a logarithmic scale between 1 and 100000 . For both the C parameter and the number of principle components the different parameters did not change the results significantly . The best parameters were 250 principle components and C = 10 . The results for the best values for the DFT window length and the time domain threshold parameters are presented in detail in the results section . We also used a five-fold cross validation approach to test for possible overfitting of the classifiers , i . e over adjustment of the classifiers to the specific training sets in a way that does not represent the actual real world data . To do this we divided the entire data ( i . e . not only the training set ) into five equal sized parts each containing a training set ( four fifth ) and a test set ( one fifth ) in the same way that was described above . For each of these five parts the entire process of finding the best parameters was executed on the training set and the performance was then tested on the relative test set . This procedure created the standard deviations of the performance measures that are presented in the results section . We used the area under the Receiver operating characteristic ( ROC ) curve to measure the performance of our classifying machines . The ROC curve is commonly used in psychophysics to estimate performance while changing a parameter . It is created by plotting the true positive rate ( TP ) on the Y axis and the false positive rate ( FP ) on the X axis , while changing a parameter . In our case the parameter along which TP and FP were plotted is the offset b of the hyperplane . Varying the offset is equivalent to moving the hyperplane along its normal direction ( in parallel to itself ) . It is obvious that on one extreme case the rate of true and false positives will both be zero , and on the other extreme they will both be 1 . Calculating the area under the ROC curves ( depict as the AUC ) evaluates the performance for all possible settings of b . The area ranges between 0 . 5–1 , where 0 . 5 means a random classifier , and 1 means a perfect one . Any other value can be interpreted as the probability of ranking a positive data point higher than a negative one in a randomly drawn pair from the test data set . The standard deviations of the performance values were calculated for the results of the five different cross validation folds . In order to compare classification performance of machines trained under different conditions ( for instance when changing one of the above parameters ) , the classification performance measures were first transformed using the arcsin transformation: ( 4 ) where P is the area under the ROC curve for a certain classification task . The transformed data was then tested for normality using both the Kolmogorov-Smirnov and the Shapiro-Wilk parametric tests , which found no significant deviations from normality in all cases . Therefore we used a two-way analysis of variance ( repeated measures ANOVA ) test to compare the classification performance , with a Tukey post hoc test in cases of more than two treatments . An alpha of α = 0 . 05 was used to determine significance . A one-way ANOVA test with a Bonferroni correction was used in the cases where the effects of a treatment were tested on the classification of a single plant species . Generating an echo from a spectrogram without phase information is impossible . In the case of our complex echoes however , the phase information is nearly random , as would be expected from a signal that is a superposition of echoes returning from many reflectors . We therefore used each column of the spectrogram as a spectrum and generated the corresponding part of the echo using a random phase . In order to prevent discontinuities when concatenating these time signals we randomly altered the phase of the frequency with the highest energy in the last created time signal such that the intensity and first derivative of its beginning matched the ones of the end of the previous time signal . This was repeated until the intensity difference was no more than 1% of the highest intensity in the last generated echo part and the first derivative of the two had the same sign . The random phase method might create problems if the spectrograms are calculated with a high overlap , because in this case the phase information in neighbouring columns is highly dependent . To verify this method and make sure that no artefacts are created , we tested whether the random phase echoes change their class membership when analysed with our trained classifiers . For the pair apple vs . corn , for which we presented the hybrid spectrograms in the Results , we trained a classifier on original spectrograms that were created with a 10% overlap between adjacent FFT windows , and used the spectrograms of the random phase echoes as a test set . Non of the echoes changed its class after the random phase manipulation , which means that our classifiers treated the random phase echoes as representing the plant species they were supposed to imitate . | Bats are able to classify plants using echolocation . They emit ultrasonic signals and can recognize the plant according to the echo returning from it . This ability assists them in many of their daily activities , like finding food sources associated with certain plants or using landmarks for navigation or homing . The echoes created by plants are highly complex signals , combining together all the reflections from the many leaves that a plant contains . Classifying plants or other complex objects is therefore considered a troublesome task and we are far from understanding how bats do it . In this work , we suggest a simple algorithm for classifying plants according to their echoes . Our algorithm is able to classify with high accuracy plant echoes created by a sonar head that simulates a typical frequency-modulated bat's emitting receiving parameters . Our results suggest that plant classification might be easier than formerly considered . It gives us some hints as to which features might be most suitable for the bats , and it opens possibilities for future behavioral experiments to compare its performance with that of the bats . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"neuroscience/behavioral",
"neuroscience",
"neuroscience/animal",
"cognition",
"neuroscience/sensory",
"systems"
] | 2008 | Plant Classification from Bat-Like Echolocation Signals |
Bacterial pathogens have evolved a specialized type III secretion system ( T3SS ) to translocate virulence effector proteins directly into eukaryotic target cells . Salmonellae deploy effectors that trigger localized actin reorganization to force their own entry into non-phagocytic host cells . Six effectors ( SipC , SipA , SopE/2 , SopB , SptP ) can individually manipulate actin dynamics at the plasma membrane , which acts as a ‘signaling hub’ during Salmonella invasion . The extent of crosstalk between these spatially coincident effectors remains unknown . Here we describe trans and cis binary entry effector interplay ( BENEFIT ) screens that systematically examine functional associations between effectors following their delivery into the host cell . The results reveal extensive ordered synergistic and antagonistic relationships and their relative potency , and illuminate an unexpectedly sophisticated signaling network evolved through longstanding pathogen–host interaction .
Many bacterial pathogens employ type III secretion systems ( T3SSs ) to deliver virulence effector proteins directly into eukaryotic host cells [1] . An essential early T3SS-dependent step in Salmonella pathogenesis is bacterial invasion of non-phagocytic intestinal epithelial cells , an event that can be modelled using cultured cells [2] . Invading bacteria deliver effectors that induce actin-rich membrane ruffles , which drive pathogen internalization into a membrane-bound vacuole where they subsequently survive and replicate [3] . Effectors are delivered into the target cell via a cholesterol-binding plasma membrane-integral translocon comprising SipB and SipC , which is likely linked to the T3SS by SipD [4] , [5] . Six delivered effectors manipulate the target cell actin cytoskeleton ( summarized in Figure 1A ) . Two Salmonella actin-binding proteins , SipC and SipA , control actin dynamics directly [6] . In addition to its role in effector delivery , discrete SipC domains nucleate actin polymerization and bundle actin filaments ( F-actin ) [7] . Both these SipC-directed activities are stimulated by SipA [8] , which itself binds and stabilizes F-actin and suppresses actin turnover by host ADF/cofilin and gelsolin [9] , [10] . Further effectors stimulate Rho-family GTPase signaling to induce cytoskeletal and nuclear responses [11] . The guanine nucleotide exchange factor mimic SopE ( or ubiquitous SopE2 ) activates Cdc42 and Rac1 GTPases directly [12]–[14] , whereas the inositol polyphosphatase SopB/SigD stimulates Cdc42 , Rac-1 and the cellular RhoG SH3-containing guanine nucleotide exchange factor ( SGEF ) through induced phosphoinositide fluxes [12] , [15] , [16] . After invasion , Rho GTPase up regulation is antagonized by SptP , a GTPase activating protein ( GAP ) mimic and tyrosine phosphatase [17] . Discrete bacterial surface proteins of Listeria and Yersinia induce internalization by hijacking host receptor-mediated endocytosis [18] . In contrast , individual Salmonella effectors are insufficient to promote bacterial internalization , although SopE , SopB and SipC do elicit generalized cell-wide cytoskeletal reorganization and membrane protrusions when expressed individually in cells [8] , [14] , [19] , [20] . It is likely that delivered effectors must therefore act in concert to induce productive actin rearrangements rapidly and specifically beneath invading Salmonellae without compromising target cell viability [6] , [11] . We have previously demonstrated that Salmonella entry effectors localize to the target cell plasma membrane both when expressed individually in cultured cells and after delivery via the bacterial T3SS , and consequently we proposed the plasma membrane as a critical interface for Salmonella effector action [19] . We next explored how the activities of these spatially co-incident effectors might potentially be coordinated to trigger actin rearrangements . Here we describe systematic experimental screens that illuminate the extent , potency and hierarchy of interplay between Salmonella effectors and their host targets within cultured cells .
To investigate the subcellular localization of effectors delivered by the invasion-associated Salmonella T3SS [19] , we generated a bank of wild-type ( WT ) S . typhimurium SL1344 strains each additionally expressing an individual plasmid-encoded epitope-tagged entry effector . This allowed the expression , secretion and delivery of an individual effector to be specifically enhanced ( 1 . 5–3 . 3-fold WT ) in a WT background ( described in detail in Materials and Methods and [19] ) . Hereafter , these WT strains engineered to deliver increased doses of particular effectors are termed ‘effector-augmented strains’ , and the enhanced effector as ‘d-effector’ . When cultured cells were infected with the effector-augmented strains ( Figure 1B , left ) , we observed that a significant increase in bacterial invasion rate occurred with the dSopE , dSopB or dSipC strains ( Figure 1B , right ∼300%; i . e . an approximately 3-fold increase compared to WT normalized to 100% ) . Each of these strains secrete the other entry effectors at WT levels [19] , but since SipC additionally functions in effector translocation in complex with SipB [7] , we additionally investigated the effect of enhancing the levels of the other translocation-associated effectors SipB and SipD ( Figure S1A ) . In contrast to the dSipC strain , no significant increase in invasion rate occurred upon infection with effector-augmented dSipB or dSipD strains ( dSipB 120±15 . 8% , dSipD 80±14%; Figure S1B ) . Nevertheless , consistent with previous observations [5] , [21] , SipB localized to both the plasma membrane and the bacterial surface , whereas SipD could only be weakly detected on the surface of some ( ∼10% ) internalized bacteria ( Figure S1C ) . Therefore , the increased invasiveness of the dSipC strain likely arises from SipC effector function rather than any unrelated effect on effector translocation . As S . typhimurium mutants lacking sptP and WT are equally invasive [17] , the observed rate increases associated with the dSopE , dSopB and dSipC strains likely reflect the stimulatory action of the augmented effector rather than decreased delivery of the signal-suppressing effector SptP . Indeed , levels of delivered SopE and SopB in WT , an sptP mutant and the dSptP strain are indistinguishable ( D . Humphreys , unpublished ) . In addition , SopE , SopB or SipC are each able to stimulate actin polymerization in cultured cells , although alone are insufficient to trigger bacterial entry [6] , [11] . Salmonella sopE or sopB deletion mutants exhibit only a modest decrease in invasion , whereas sipC null mutants are non-invasive , as SipC is not only an effector but also a translocator of other effectors [4] , [16] , [22] . These combined data therefore strongly support linkage between the delivered dose of stimulatory effectors , the resulting extent of actin polymerization and Salmonella invasion rate . However , when cultured cells were transfected with effectors ( denoted ‘t-effector’ ) and infected with WT S . typhimurium ( Figure 1C , left ) , only mild rate increases ( ∼120% i . e . an approximately 0 . 2-fold increase when compared to WT normalized to 100% ) were evident . A comparable effect was observed when control Cdc42-dependent actin polymerization and macropinocytosis were induced with bradykinin ( Figure 1C , right ) . tSptP and tSipB conferred a limited resistance to invasion ( 54±37% and 56±22% , respectively ) , consistent with their ability to counteract cytoskeletal and nuclear responses and to disrupt cellular endomembranes , respectively ( Figure 1C , right; Figure S1D; Figure S1E; [17] , [23] ) . Additionally , we verified that transfectants remained viable and that effectors were expressed and localized correctly ( Figure S2A ) . Indeed , the concentration of each t-effector in the plasma membrane exceeded that of the corresponding d-effectors in infected cells ( Figure S2B ) . Thus , although tSopE , tSopB and tSipC all trigger dramatic cytoskeletal reorganization [19] , unlike delivered effectors ( Figure 1B , right ) , these transfected effectors are unable to significantly influence invasion rate , either because the effectors distribute along the plasma membrane throughout the cell , rather than at specific pathogen contact sites as with the T3SS-delivered effectors , or because their premature activity inhibits WT entry . Together these data reveal that the spatio-temporal context of effector activity seems critical , as only cytoskeletal rearrangements induced by delivered stimulatory effectors can significantly increase bacterial invasion rate . To determine whether the enhanced invasion rates of the dSopE , dSopB and dSipC strains reflect specific effector-induced stimulation of defined cellular signaling pathways , we next investigated the consequence of transfecting cells with dominant negative Rho-family GTPases prior to infection with WT Salmonella or each of the effector-augmented strains ( Figure 2A ) . As expected , tCdc42 ( N17 ) , tRac1 ( N17 ) and tRhoA ( N19 ) expression suppressed WT S . typhimurium invasion by ∼40% , confirming Rho-family GTPases as important cellular targets of delivered Salmonella effectors ( Figure 2B; [14] , [24] ) . When tCdc42 ( N17 ) or tRac1 ( N17 ) transfectants were infected with the dSopE or dSopB strains , the previously observed increases in invasion rate significantly diminished ( e . g . tCdc42 ( N17 ) :dSopE –163%; tRac1 ( N17 ) −238% ) , yet remained unaltered by tRhoA ( N19 ) expression . Conversely , none of the expressed dominant negatives suppressed the enhanced invasion rate of the dSipC strain ( e . g . tCdc42 ( N17 ) +19%; tRhoA ( N19 ) +13% ) , indicating that SipC activity is Rho-family GTPase independent ( Figure 2B ) . Furthermore , unlike the WT , the invasion rate of the dSipC strain was equivalent in each dominant negative background ( i . e . in Figure 2B , compare constant dSipC invasion rate to decreases in WT rate in the dominant negative backgrounds ) , suggesting that augmenting SipC delivery can compensate for the reduction in SopE/SopB-dependent stimulation of Cdc42/Rac1-dependent signaling . This is consistent with the ability of SipC to induce actin reorganization directly [7] . These data illustrate that three entry effectors cooperate in the WT strain to stimulate parallel Cdc42/Rac1-dependent ( SopE , SopB ) and –independent ( SipC ) invasion pathways , the relative contribution of which can be modulated by the comparative levels of each effector delivered . Based on this transfection-infection approach , we next designed binary entry effector interplay ( BENEFIT ) screens to assess potential cross talk between pairs of effectors . In the first ‘trans’ BENEFIT screen , we investigated the effects of expressing individual bacterial effectors rather than dominant negative derivatives of host proteins in cells prior to infection with WT Salmonella or the effector-augmented strains ( Figure 3A ) . We conservatively defined functional interplay between effector A ( ‘d-effector’; augmented levels of which are delivered via the T3SS in a WT background ) and effector B ( ‘t-effector’ pre-expressed in the target cell by transient transfection ) to be an increase ( or decrease ) in rate of at least one-fold of WT after correction for WT Salmonella invasion of cells expressing effector B ( Figure 1C , right ) , and invasion of control cells by WT Salmonella expressing augmented levels of effector A ( Figure 1B , right ) . This threshold is significantly different from the controls ( Mann Whitney U p<0 . 05 ) . The results of this screen were striking , revealing reproducible changes in Salmonella invasion rates ( Figure 3B; Figure S3 ) . Our criteria identified four synergistic relationships . Two effector combinations induced mild increases in invasion rates ( dSopE:tSipC +128%; dSopB:tSopE +150% ) , whereas two further pairings stimulated even more prominent changes ( dSipC:tSipA +369%; dSopB:tSipC +398% i . e . the invasion rate was ∼4-fold greater than WT entry into control cells ) . Direct observation of infected cells in parallel revealed dramatic actin reorganization induced by these pairings ( Figure S4 ) , the morphology of which often reflected a combination of their reported activities [8] , [14] , [19] . These rate and phenotypic variations were not a simple function of augmenting effector dose , as no significant effects were observed using homotypic effector combinations under identical conditions ( i . e . dSopE:tSopE +8%; dSipC:tSipC +22%; Figure 3B and Figure S2C ) . Furthermore , all these effects were apparently also dependent on ordered effector activity , as inverse pairing resulted in inhibition of bacterial invasion ( e . g . dSipC:tSopE −76%; dSipA:tSipC –120% ) . An extended screen including the dSipB and dSipD strains confirmed that unlike SipC , which is involved in both effector delivery and actin reorganization [4] , [7] , [22] , SipB and SipD neither functionally contribute to invasion nor engage in synergy with any delivered effectors ( Figure S3A; Figure S3B ) . Significant inhibitory effects were also evident ( Figure 3B ) . Premature expression of either SopB or SptP in cells suppressed subsequent infection by several Salmonella strains , in particular dSopE ( tSopB , −196%; tSptP , −159% ) , dSopB ( −127% , −105% ) or dSipC ( −213% , −123% ) , and as with effector synergy , this functional antagonism was reversed ( e . g . tSipC:dSopB +398%; tSopE:dSopB +150% ) or annulled ( e . g . tSipC:dSptP −11%; tSopE:dSptP +16% ) using inverse pairings . Notably , although both tSptP and tSipB impede WT invasion ( tSptP 54±37%; tSipB 56±22% ) , the effector-augmented strains were not significantly inhibited by tSipB expression ( Figure S3B ) . The inhibitory effects reflect true variations in invasion rate rather than indiscriminate changes due to differential effector-induced cellular toxicity , as effector transfection generated no significant alteration in the number or viability of target cells ( Figure S2A ) , a conclusion reinforced further by the direct observation of internalized bacteria ( Figure S4 ) . While trans BENEFIT screening permitted the sequence of effector activity to be dictated experimentally , an associated limitation is the imposed effector concentration imbalance , as the effector levels in transfected cells exceed those delivered by the bacteria . However , when effector concentrations were assessed following infection and transfection by quantitative immunoblotting , analysis revealed the difference between transfected and delivered effector concentrations was less than one order of magnitude ( 3–8 fold excess in transfectants; Figure S2B ) . Fractionation of transfected cells following infection with the effector-augmented strains also demonstrated that both the transfected and delivered effectors remained co-localized in the plasma membrane and cytoskeletal fractions , as when expressed or delivered in isolation ( Figure S2C; [19] ) . Coupled to the fact that a ‘dominant negative’ phenotype is induced only by tSopB and tSptP , these control data further suggest that disordered effector activity rather than non-specific effector concentration or mislocalization likely accounts for this antagonism . Thus , the trans BENEFIT data suggest that extensive functional interplay does occur between delivered Salmonella effectors . We next modified the BENEFIT screen to eliminate the imposed ordered effector activity and the concentration bias resulting from transfection . In this ‘cis’ screen , cultured cells were infected with pairwise mixtures of effector-augmented strains , in combination with each other or WT S . typhimurium ( Figure 4A ) . At first , an identical multiplicity of infection to the trans screen was used , except that each infection mixture comprised an equal proportion of two distinct effector-augmented S . typhimurium strains . An antibiotic marker was introduced into one strain to allow the relative invasion efficiency of each to subsequently be calculated . Since the selection conferred a marginal competitive disadvantage during mixed infection when compared to the isogenic non-resistant strain ( ∼−5% ) , each infection was performed in duplicate , allowing the strain carrying the marker to be alternated . The mean invasion efficiency of each strain could then be calculated and controlled for the marker-induced variation ( Figure 4A ) . We initially confirmed that 50∶50 mixtures of isogenic input strains were recovered at an equivalent output ratio , i . e . strains that differed only with respect to the resistance marker did not compete with each other and were recovered with equivalent frequency from the infected cells [Figure 4B; e . g . mixed WT:WT invasion rate was 100% to which component strains contribute equivalently ( 50%∶50% ) ] , and that the enhanced invasion rates observed with the dSopE , dSopB and dSipC strains ( Figure 1B , right ) were recapitulated during mixed infection [Figure 4B; e . g . the invasion rate of a mixed dSopE:dSopE strain input remains ∼3-fold WT ( 304% ) to which component strains contribute equivalently ( 152%∶152% ) ] . Having experimentally verified these important assumptions , we next performed a complete cis BENEFIT screen using heterologous mixtures of effector-augmented WT strains . As previously , we conservatively defined functional interplay between effector A ( ‘d-effector’ ) and effector B ( ‘d-effector’; delivered by a separate strain ) to be reflected by an increase ( or decrease ) in invasion rate of at least one-fold of WT from the change recorded when WT Salmonella expressing augmented levels of either effector A or B enter cells independently ( Figure 1B , right ) . Such changes are statistically significantly different from the controls ( Mann Whitney U p<0 . 05 ) . As might be expected , the resulting variations in invasion rate were lower in magnitude than in the trans BENEFIT screen , again illustrating that effector concentration is related to invasion rate , provided that host targets ( e . g . actin , Rho GTPases , phosphoinositides ) remain in excess . Two significant synergistic relationships were immediately evident ( Figure 4B , dSipA:dSipC , +227%; dSipC:dSopB , +169% ) that independently corroborated the most significant data from the trans BENEFIT screen ( Figure 3; Figure S3 ) , and an additional potential synergy also emerged , albeit nearer to the significance threshold ( Figure 4B , dSipA:dSopB , +129% ) . Distinct synergistic classes could also be defined; ‘mutual’ where both strains benefit from their association [e . g . both the dSipC ( +74% ) and the dSipA ( +153% ) strains gain from a mixed dSipC:dSipA infection] , or ‘selfish’ in which one strain exploits its partner [e . g . only the dSipA strain ( +144% ) benefits from a mixed dSipA:dSopB infection , whereas invasion of the dSopB strain is unchanged or marginally compromised ( −16% ) ] . In contrast to the trans BENEFIT screen , no significant antagonistic effects were observed ( Figure 4B ) . This is consistent with the view that inhibition arises from transfecting the cells with effectors prior to invasion that subsequently interfere with the ordered activities of bacterially-delivered effectors . Co-infection with the effector-augmented strains never enhanced WT S . typhimurium invasion . Indeed , when mixed with the dSopE or dSipA strains the WT was appreciably but not significantly disadvantaged , as in mixed WT:dSipA and WT:dSopE infections , the WT invasion rate was reduced from the expected 50% to 11% and 26% , respectively ( Figure 4B ) . Enhanced expression of SipC also conferred a mild selective advantage over the WT , as invasion of the dSipC strain increased from the expected 112% to 160% when co-infected with WT ( Figure 4B ) . These data illustrate that one strain does not simply passively assist the invasion of any partner . Rather , the increased invasion conferred on any particular strain relates to the concentration and activity of the effectors within the target cell delivered by the strain itself or its co-infecting partner . To confirm this further , we performed additional assays in which the initial infection mixtures were artificially biased ( 90:10/10:90 ) to favour one or other of the strains . Under these conditions , the previously observed synergy was abolished as the recovered output reproducibly mirrored the composition of the input mixture ( Figure S5 ) . A threshold level of both delivered effectors is therefore necessary to drive the observed synergy . Our BENEFIT screening suggested that unexpectedly sophisticated interplay occurs between multiple delivered Salmonella entry effectors . Next , we further investigated one of the previously unrecognized relationships using complementary genetic , biochemical and cell biology techniques . We selected SipC-SopB synergy for further analysis , as this was apparently a dominant association identified by both the trans and cis BENEFIT screens . Initially , we wished to establish whether known SipC and SopB activities were required for synergy . The SipC N-terminal domain ( SipC-N; residues 1–120 ) directs actin filament bundling , whereas the C-terminal domain ( SipC-C; 200–409 ) induces actin nucleation [7] . Correspondingly , tSipC-N and tSipC-C both reorganized actin when expressed in cultured cells ( Figure 5A ) . tSipC-C localized to the leading cell edge where it induced lamellipodial and filopodial protrusions reminiscent of those generated by membrane-integral SipC [8] , [19] , yet was distributed between the cytoskeletal , cytosolic and plasma membrane fractions upon mechanical fractionation ( Figure 5A , SipC-C ) . tSipC-N localized along induced hyper-elongated filopodia 30–60 µm in length , an activity similar to the cellular actin-bundling protein fascin [25] , and predominantly to the plasma membrane fraction upon mechanical fractionation ( Figure 5A , SipC-N ) . However , when cells expressing SipC-N or SipC-C were infected with effector-augmented S . typhimurium strains , SipC-SopB synergy was attenuated ( Figure 5B ) . Indeed , unlike tSipC , tSipC-N significantly inhibited invasion by the dSopE , dSopB and dSipC strains ( Figure 5B , i . e . compare dSopE:tSipC-N , −248%; dSopB:tSipC-N , −168%; dSipC:tSipC-N , −236% to dSopE:tSipC , +128% , dSopB:tSipC , +398% , dSipC:tSipC , +22% ) , as premature actin bundling is likely to intrinsically impede cytoskeletal plasticity required for bacterial invasion . tSipC-C also inhibited dSopE and dSipC invasion ( Figure 5B , dSopE:tSipC-C , −273%; dSipC;tSipC-C , −129% ) , but could still apparently engage in limited synergy with the dSopB strain , albeit below our imposed significance threshold ( dSopB:tSipC-C , +85% ) . These data suggest a potential link to C-terminal domain function , but additionally imply that plasma membrane localization of this region is important for efficient synergy with SopB , reaffirming the view that the plasma membrane is a critical interface for effector interplay [19] . Next , we investigated whether SopB inositol phosphatase activity was required for SipC-SopB synergy by examining the effects of expressing the homologous invasion-associated inositol phosphatase Shigella flexneri IpgD [26] or a phosphatase-dead SopB derivative containing a C462S mutation in the active site [SopBC462S; [20]] prior to infection with the bank of effector-augmented S . typhimurium strains ( Figure 6A ) . When pre-expressed in cells , tSopB is a ‘dominant-negative’ inhibitor of the dSipA , dSipC , dSopE and dSopB strains ( Figure 3B; Figure 6B ) . This effect was recapitulated using tIpgD ( dSipA:tIpgD , −53%; dSipC:tIpgD , −106%; dSopE:tIpgD , −231%; dSopB:tIpgD , −76% ) , although inhibition of the dSipA and dSopB strains was somewhat reduced . Nevertheless , this functional substitution suggested that inhibition is driven by inositol phosphatase activity rather than an unknown SopB-specific function . In clear support of this , significant inhibition was alleviated using phosphatase-dead tSopBC462S ( Figure 6B; dSipA:tSopBC462S , −11%; dSopE:tSopBC462S , −95%; dSopB:tSopBC462S , −42%; dSptP:tSopBC462S , −8% ) , and invasion of the dSipC strain specifically but mildly enhanced ( dSipC:tSopBC462S , +97% ) . To investigate the effect of SopB inositol phosphatase activity on SipC synergy during infection , we generated a S . typhimurium strain engineered to deliver increased levels of SopBC462S in the WT background ( dSopBC462S ) , and included this in a trans BENEFIT screen where effector-transfected cells were infected with dSopE , dSopB or dSopBC462S strains ( Figure 7A ) . Synergy between tSipC and dSopBC462S was markedly attenuated to near the statistical threshold ( Figure 7B , compare tSipC:dSopB , +298% to tSipC:dSopBC462S , +111% ) , whereas the weaker tSopE synergy was abolished ( Figure 7B , compare tSopE:dSopB , +150% to tSopE:dSopBC462S , +19% ) . The latter is a recognized relationship mediated by cellular inositol phosphates [16] . Additionally , unlike the dSopB strain , dSopBC462S invasion was not inhibited in cells expressing SipA , SopB or SptP ( e . g . compare tSptP:dSopB , −105% to tSptP:dSopBC462S , −25% ) . These combined findings suggested that SopB inositol phosphatase activity contributes to synergy with membrane localized SipC . SopB additionally aids invasion by cleaving plasma membrane phosphatidylinositol-4 , 5-bisphosphate [PI ( 4 , 5 ) P2] to promote membrane elasticity and vacuole formation [20] . To establish whether levels of PI ( 4 , 5 ) P2 influence invasion the augmented dSopB strain , we expressed phosphatidylinositol-4-phosphate-5-kinase [PIP ( 5 ) K] , which catalyzes the formation of PI ( 4 , 5 ) P2 , in cells prior to infection with the effector augmented S . typhimurium strains . Strikingly , excess PI ( 4 , 5 ) P2 stimulated WT Salmonella invasion more than two-fold ( Figure 8A ) , and specifically and significantly enhanced dSopB strain invasion ( Figure 8A; tPIP ( 5 ) K:dSopB , +608% ) , while inhibiting the dSopE strain ( Figure 8A; tPIP ( 5 ) K:dSopE , −320% ) . The phenotypic similarity between tPIP ( 5 ) K:dSopB and tSipC:dSopB suggested that SipC might generate PI ( 4 , 5 ) P2 either directly or via PIP ( 5 ) K stimulation . Indeed , the periphery of Salmonella-induced membrane ruffles are enriched for PI ( 4 , 5 ) P2 [20] . To pursue this hypothesis , we assessed the relative proportion of phosphatidylinositol species in cells expressing Salmonella effectors by radioactive counting of fractions separated by ion exchange chromatography ( Figure 8B ) . When compared to resting cells , tSopB lysates contained significantly enhanced levels of phosphatidylinositol-3 , 4 , 5-trisphosphate [PI ( 3 , 4 , 5 ) P3] and phosphatidylinositol-3 , 4-bisphosphate [PI ( 3 , 4 ) P2] , likely to be a stable breakdown product of PI ( 3 , 4 , 5 ) P3 [27] , and to a lesser extent phosphatidylinositol-3-phosphate [PI ( 3 ) P] ( Figure 8B , tSopB ) . By comparison , no major alterations were observed when comparable tSipC or tSptP lysates were assayed ( Figure 8B , tSipC and tSptP ) . These biochemical data demonstrate that SipC itself does not directly or indirectly generate PI ( 4 , 5 ) P2 , but do not preclude that SipC alters the local concentration or distribution of existing PI ( 4 , 5 ) P2 in the membrane . To examine this possibility , the PI ( 4 , 5 ) P2-binding plextrin-homology ( PH ) domain of phospholipase Cδ fused to green fluorescent protein ( PLCδ-PH-GFP ) was expressed in cultured cells prior to infection with effector-augmented strains to report PI ( 4 , 5 ) P2 distribution during invasion [20] . As expected , the membrane ruffles associated with actin rearrangements induced by the dSopB strain were largely devoid of PI ( 4 , 5 ) P2 , indicated by the diffuse reporter probe distribution ( Figure 8C , dSopB PLCδ-PH-GFP ) . In contrast , membrane ruffles generated by the dSipC strain were enriched with PLCδ-PH-GFP , which was frequently coincident with induced actin rearrangements and concentrated at filopodial tips ( Figure 8C , dSipC ) . Clusters of PLCδ-PH-GFP were also clearly evident at the plasma membrane , a distribution strikingly similar to that of SipC itself [8] , [19] . Although cells infected with the dSopE strain also induced profuse membrane ruffles morphologically more reminiscent of the dSipC strain [19] , no PI ( 4 , 5 ) P2 enrichment or clustering was evident ( Figure 8C , dSopE ) . These data suggest that localized accumulation or clustering of plasma membrane PI ( 4 , 5 ) P2 by SipC enhances the availability of SopB substrate at bacterial entry foci . These findings from biochemical and cell biology approaches demonstrate the physiological context of a novel relationship highlighted by our genetic BENEFIT screening and extend our mechanistic understanding of SipC-SopB interplay .
Emerging evidence has inferred that multiple effectors delivered by bacterial T3SSs act cooperatively within the target cell to subvert host signaling processes [8] , [17] , [28] . Salmonella invasion of host cells is a paradigm system in which the biochemical activities of the central T3SS effectors that manipulate cellular actin dynamics are likely identified [6] , [11] . However , the nature and extent of cross talk between these effectors that co-localize at the target cell plasma membrane remained unknown [19] . Our findings provide initial insights into a sophisticated program of ordered effector activities underlying cell invasion by Salmonella ( Figure 9 ) . The BENEFIT screening data identify extensive novel Sip-Sop synergy ( SipC-SopB; SopB-SipA; SipC-SopE ) , and additionally validate every cooperative [SipA-SipC [8]; SopB-SopE [16]] and antagonistic [SopE-SptP [17]] effector relationship detected by previous genetic or biochemical approaches . We demonstrate that SipC promotes actin polymerization independently of Rho-family GTPases . This is clearly consistent with the ability of SipC to nucleate actin polymerization directly [7] , and support a fundamental role for this function during invasion [22] , [29] . SipC also emerges as a central participant in a more extensive signaling network , as it engages in potent synergistic relationships with delivered SipA and SopB , and to a lesser extent with SopE . Such early intracellular communication between SipC and Sops reconciles the observation that triple Salmonella mutants lacking sopE , sopE2 and sopB are unable to efficiently induce actin rearrangements [16] . Network architecture implies that SipC not only cooperates with SipA to directly trigger actin reorganization [8] , but also concurrently reinforces signaling by indirectly stimulating SopB and SopE activity . This ‘feed-forward’ amplification and evolved effector interdependency may allow the Cdc42/Rac1-dependent pathway to dominate during WT infection [24] . Our data reveal an important role for phosphoinositides in mediating the dominant and previously unrecognized synergy between SipC and SopB . PI ( 4 , 5 ) P2 clusters in membrane ruffles generated specifically by SipC , and is subsequently hydrolysed by SopB . As only the dSopB strain gains from co-infection with the dSipC strain , this suggests that prior PI ( 4 , 5 ) P2 concentration at entry sites either by dSipC or experimentally with tSipC subsequently enhances the invasion of the dSopB strain , which is more adept at ruffle closure by virtue of augmented inositol phosphatase activity . Like SipC , we show that PI ( 4 , 5 ) P2 concentrates locally at the tip of induced filopodia [19] . This is consistent with co-secreted cholesterol-binding SipB initially integrating into cholesterol-rich regions of the plasma membrane [5] , where PI ( 4 , 5 ) P2 also concentrates . It is also tempting to speculate that PI ( 4 , 5 ) P2 might also influence bacterial SipC actin nucleation activity , as this species is a stimulatory co-factor of cellular Arp2/3-dependent actin polymerization [30] , [31] . Similarly , our findings that tSopB generates PI ( 3 , 4 , 5 ) P3 , PI ( 3 , 4 ) P2 and PI ( 3 ) P concurs with reports of SopB-dependent stimulation of RhoG via SGEF and PI ( 3 ) P accumulation on Salmonella-containing vacuoles [12] , [32] . BENEFIT screening not only defines such previously undetected relationships between delivered effectors , but also supports the view that a controlled programme of temporal effector activities drives Salmonella invasion . In the trans screen , the detected synergies are unidirectional , i . e . tSipC enhances invasion of the dSopB strain but not vice versa , an effect recapitulated in the cis screen , when co-infecting strains synergise selfishly or mutually . Also , tSopB or tSptP expression in cells prior to infection selectively inhibits invasion of particular augmented strains . This experimentally imposed asynchronous SopB or SptP activity likely inhibits subsequently delivered effectors by sequestering or down-regulating essential cellular targets ( Rho family GTPases ) or by prematurely inter-converting necessary secondary messengers through SopB inositol phosphatase activity or SptP-dependent protein dephosphorylation . As tSptP also reduces WT invasion , this might reflect a premature generalized GAP-driven protective effect [17] . However , tSopB only inhibits the invasion of certain effector-augmented strains ( i . e . not the WT ) indicative of specific functional interplay , a phenotype abolished when tSopB derivatives are expressed that lack inositol phosphatase activity . The BENEFIT screens do not distinguish whether these temporal relationships reflect the ordered activity of simultaneously delivered effectors or a requirement for sequential effector delivery via the T3SS . The available experimental data from WT infection still remain consistent with either scenario: delivered effectors co-localize at the plasma membrane [19]; ‘early’ acting SopE and ‘late’ SptP are present in cells at equimolar concentrations after 15 minutes [33]; and imaging of effector delivery suggested equivalent kinetics for both SipA and SopE transfer [34] . What is clear from our data is that the T3SS has evolved to ensure that delivered effectors are precisely dosed in the WT to balance efficient bacterial invasion and host cell viability . Although most effector relationships are common to both BENEFIT screens , SipA-SopB synergy that favoured the dSopB strain emerged uniquely from the cis screen . Why this phenotype was constrained to mixed infection is as yet unclear , although it might relate to the recently identified role of these effectors in invasion-associated tight junction disruption [35] . Provocatively , both these effectors also have additional intracellular roles in controlling the subsequent maturation and positioning of Salmonella containing vacuoles [32] , [36] , although whether these functions are simultaneously activated 60 minutes after infection remains unknown . Our data provide the first systematic experimental study examining the scope , nature and relative potency of interplay between bacterial effectors delivered into eukaryotic cells . Nevertheless , as with any experimental system BENEFIT screens inevitably also have associated limitations . These include the use of a cultured cell model and genetically modified Salmonella strains , and the possibility that changes in invasion rate used as a readout result from temporal alterations or asynchrony in the process or differential effector stability rather than interplay . These aspects are technically challenging to monitor and control . However , coupled to the fact that our screening detected every previously identified effector association , our extended investigation of the novel SipC-SopB synergy using complementary biochemical and cell biology approaches , provides strong supporting evidence that the previously unrecognized relationships emerging from our screening reflect bona fide effects . Nevertheless , each of the identified associations now requires further in depth investigation . Biochemical and structural studies have illustrated that many effectors comprise defined domains that act as functional modules [37] , for example SipA contains a C-terminal actin-binding domain and an N-terminal region that influences phagosome positioning [9] , [36] . As we have demonstrated with our analysis of SipC-C and SipC-N function and the catalytic-dead SopB derivative , BENEFIT screening assays can now be exploited further to decipher the relative contribution of such discrete functions to effector interplay , and to examine how as yet undefined Salmonella effectors might contribute to invasion . As demonstrated using Rho-family GTPases as proof-of-principle , dominant negative and RNAi approaches can also be employed to examine the involvement of additional host targets in distinct effector-stimulated signaling pathways . The assays could also be adapted to examine virulence proteins in other pathogen-induced processes , provided that there is a readily quantifiable output phenotype such as bacterial replication or dissemination .
Construction of S . typhimurium SL1344 strains engineered to express augmented levels of individual entry effectors in a WT background has been described in detail previously [19] . Briefly , S . typhimurium SL1344 was transformed with expression plasmids pB:SopEFLAG , pB:SopBFLAG , pB:SopBFLAG pB:SipA , pB:SptPFLAG , pB:SipC , pB:SipB or pB:SipD generated by PCR amplification of sopE , sopB , sipA , sptP , sipC , sipB or sipD engineered to contain NdeI and HindIII sites from an SL1344 chromosomal DNA template , into pTrc99A-FF4 to allow constitutive low-level expression downstream of the Trc promoter [38] , [39] . pB:SopBC462SFLAG was generated identically , except the TGT codon at nucleotide 1383 of sopB was mutated to TCC to generate a cysteine to serine substitution at amino acid 462 . Immunoblotting and densitometry of cell lysates and secreted proteins from each strain showed that plasmid-encoded SopE , SopB , SipA , SptP or SipC expression was specifically but mildly augmented ( 1 . 5–3 . 3±0 . 28 fold WT ) due to the low copy number plasmid in the absence of induction , and that this elevated expression correlated with increased effector secretion ( 1 . 5–2 . 0±0 . 28 fold WT ) . This also correlated with a mild increase in concentrations of delivered effectors , as each could be detected by immunofluorescence or immunoblotting of cell lysates after infection with the engineered strains but were below the detection threshold after WT infection under the same conditions [19] . Exogenous expression of plasmid-encoded effectors did not interfere with the expression , secretion or delivery of other chromosomal SPI1- or SPI2-secreted effectors [[19] , and L . Brawn , unpublished observations] . Expression and secretion of SipB and SipD are documented in Figure S1 . S . typhimurium SL1344 strains were maintained on Luria-Bertani agar or cultured in tryptone yeast medium ( 2TY ) containing ampicillin or spectinomycin ( 50 µgml−1 ) , as appropriate . NIH3T3 fibroblasts were cultured in Dulbecco's modified Eagle's medium ( DMEM ) supplemented with 10% ( v/v ) fetal calf serum ( FCS ) , L-glutamine and penicillin/streptomycin ( Sigma/Invitrogen ) ( 37°C , 5% CO2 ) . When required cells were treated with bradykinin ( 30 ngml−1; 37°C , 5% CO2 , 60 min ) . To assess number and viability , ∼1×106 cells were scraped into PBS , ice cold 70% ( v/v ) ethanol added ( 10vol ) , and the mixture incubated ( 4°C , 30 min ) . Cells were then pelleted ( 900g , 5 min ) and resuspended in PBS . RNAse and propidium iodide were added , and the mixture incubated ( 37°C , 30 min ) . Cell clumps were dispersed by passage through a 25-gauge needle and samples analysed for forward scatter , side scatter and fluorescence on a FACScan flow cytometer ( Becton Dickinson , USA ) , collecting >20000 events per sample . Eukaryotic expression vectors pT:SopEFLAG , pT:SopBFLAG and pT:SptPFLAG containing sopE , sopB , and sptP were generated by PCR amplification from the S . typhimurium SL1344 chromosome and cloning of products , engineered to contain BamHI and XhoI sites into pT:FLAG [19] , a fusion of AlwNI-digested phrGFP-NUC and pIRES-hrGFP ( Stratagene ) . pT:SipA , pT:SipC and pT:SipB were constructed by cloning of sipA , sipC or sipB PCR products engineered to contain XbaI and HindIII sites into pcDNA3 . 1 ( Invitrogen ) . pT:SipCNFLAG and pT:SipCCFLAG were generated by PCR amplification of nucleotides 1–360 and 600–1230 of sipC corresponding to amino acids 1–120 and 200–409 and cloning into pT:FLAG . pT:IpgDFLAG was generated by PCR amplification of ipgD from the Shigella flexneri M90T virulence plasmid and cloning into pT:FLAG . pT:SopBC462SFLAG was generated as described for pB:SopBC462SFLAG , except that the PCR product was cloned into pT:FLAG . pPLCδ-PH-GFP expressing the PH-domain of phospholipase Cδ that specifically binds phosphatidylinositol-4 , 5-bisphosphate [PI ( 4 , 5 ) P2] fused to C-terminal GFP was a kind gift from Phill Hawkins ( Babraham Institute , Cambridge ) . pT:PIP ( 5 ) K was generated by PCR amplification of the cDNA encoding murine phosphatidylinositol-4-phosphate-5-kinase and cloning of the product into pcDNA3 . 1 . The ‘cis’ or ‘trans’ binary entry effector interplay ( BENEFIT ) screens are a modification of the gentamicin protection assay , which measures the ability of Salmonella strains to invade cultured cells [40] . Stationary phase S . typhimurium SL1344 cultures transformed with pB:sopEFLAG , pB:sopBFLAG , pB:sptPFLAG , pB:sipCFLAG , pB:sipA , pB:sipB , pB:sipD or vector control [19] , were diluted 1∶500 in 2TY and incubated to maximise invasion efficiency ( 6 h , 37°C , 225 r . p . m . ) . Bacteria were washed and resuspended in DMEM supplemented with L-glutamine , and added at a multiplicity of infection ( MOI ) of 50 to 2×104 serum-starved NIH3T3 fibroblasts , transfected 48 h previously with expression plasmids pT:sopEFLAG , pT:sopBFLAG , pT:sopBC462SFLAG , pT:sptPFLAG , pT:sipCFLAG , pT:sipCNFLAG , pT:sipCCFLAG , pT:sipA , pT:ipgDFLAG , pT:PIP ( 5 ) K or empty vector [19] using Lipofectamine™ , according to the manufacturer's instructions ( Invitrogen ) . Transfection efficiency , assessed by GFP-NLS co-expression or immunofluorescence [19] was always >85% . After incubation ( 37°C , 5% CO2 , 60 min ) , cells were repeatedly washed with warm PBS and extracellular bacteria killed with gentamicin ( 100 µgml−1 in DMEM; 37°C , 5% CO2 , 60 min ) . Cells were washed again with PBS and lysed in 10 mM Tris-Cl , pH 7 . 4 , 0 . 5% ( v/v ) triton X-100 . Serial lysate dilutions were plated onto LB agar and the percentage of intracellular bacteria compared to the original inoculum . Rates were referenced against invasion of WT S . typhimurium SL1344 and an isogenic invasion deficient mutant lacking invG , an essential structural component of the T3SS , assigned as 100% and 0% respectively , in each transfectant background . Functional interplay between effector A ( ‘deffector’; delivered via the T3SS ) and effector B ( ‘teffector’ pre-expressed in the target cell by transient transfection ) was defined as an increase ( or decrease ) in rate of at least one-fold ( i . e . ±100% ) of WT after correction for WT Salmonella invasion of cells expressing effector B and invasion of control cells by WT Salmonella expressing augmented levels of effector A . Stationary phase S . typhimurium SL1344 cultures transformed with pB:sopEFLAG , pB:sopBFLAG , pB:sptPFLAG , pB:sipCFLAG , pB:sipA or empty vector [19] , were diluted 1∶500 in 2TY and incubated to maximise invasion efficiency ( 6 h , 37°C , 225 r . p . m . ) . Bacteria were washed and resuspended in DMEM supplemented with L-glutamine and mixed at a 50∶50 or 90∶10/10∶90 ratio as appropriate to give an MOI of 50 , when 2×104 serum-starved NIH3T3 fibroblasts were infected . An antibiotic marker ( spectinomycin ) was introduced into one strain to allow the relative invasion efficiency of each strain to subsequently be calculated . Since selection conferred a marginal disadvantage ( ∼5% ) , each infection was performed in duplicate to allow the marker to be alternated . After incubation ( 37°C , 5% CO2 , 60 min ) , cells were repeatedly washed with warm PBS and extracellular bacteria killed with gentamicin ( 100 µgml−1 in DMEM; 37°C , 5% CO2 , 60 min ) . Cells were washed again with PBS and lysed in 10 mM Tris-Cl pH 7 . 4 , 0 . 5% ( v/v ) triton X-100 . Serial dilutions were plated onto LB agar with and without spectinomycin and the percentage of intracellular bacteria compared to the original inoculum . In the cis assay , functional interplay between effector A ( ‘dA’; delivered via the T3SS ) and effector B ( ‘dB’; delivered by the T3SS of a separate strain ) was defined as an increase ( or decrease ) in rate of at least one-fold ( i . e . ±100% ) of WT after correction for the rate expected when either strain invades cells independently . In both screens , the results presented are the mean±SEM of 4 independent experiments each performed in triplicate and the changes reported statistically significant from the controls ( Mann Whitney U p<0 . 05 ) . Infected or transfected cells were washed in PBS and fixed with 3 . 7% ( v/v ) paraformaldehyde for fluorescence microscopy . Fixed samples were permeabilized with 0 . 2% ( v/v ) triton X-100 in PBS , blocked with 3% ( w/v ) bovine serum albumin ( BSA ) in PBS ( 1 h , RT ) , then incubated with appropriately diluted primary antibodies ( anti-FLAG monoclonal or anti-effector polyclonal ) in PBS ( 1h , RT ) . Samples were sequentially incubated with AlexaFluor 488-conjugated anti-rabbit or anti-mouse IgG secondary antibodies , according to manufacturer's instructions ( 30 min , RT; Invitrogen ) , then Texas Red-conjugated phalloidin ( 20 min , RT; Invitrogen ) and 4′ , 6′-diamidino-2-phenylindole dihydrochloride ( DAPI , Sigma ) in PBS . Coverslips were mounted using ProLong Anti-fade reagent ( Invitrogen ) , and visualized using a fluorescence microscope ( Leica DM IRBE ) . Images were captured using a CCD digital camera ( Hamamatsu ) and processed using OpenLab software ( Improvision ) . Where appropriate GFP fluorescence was observed directly in fixed cells . Bacterial DAPI signal was artificially enhanced to normalise the intensity to that of nuclear staining . Effector protein concentrations in transfected cells or delivered into cells ( measured 60 min after S . typhimurium infection ) were assayed by immunoblotting mechanical cell lysates or subcellular fractions with antibodies raised against purified effectors [19] . Signal intensities were determined using NIH Image freeware ( http://rsb . info . nih . gov/nih-image/ ) , and intensity values compared to those derived by immunoblotting known concentrations of purified effectors . Cells were fractionated as previously described [19] . Analysis was carried out as described [41] . Briefly , transfected NIH3T3 cells were washed and incubated with phosphate-free DMEM supplemented with 2% ( v/v ) dialysed heat-inactivated FCS , 0 . 2% ( w/v ) fatty acid-free BSA and ∼0 . 3 mCi [32P]Pi ml−1 in 10 mM HEPES pH7 . 4 ( diluted from an iso-osmotic stock ) to allow labelling of cellular phosphoinositides ( 37°C , 5% CO2 , 75 min ) . Cells were rapidly aspirated and fixed with 1 vol ice-cold 1 . 0 M HCl , and scraped into glass vials containing 4 vol CHCl3:MeOH ( 2∶1 , v/v ) , Folch carrier lipid and tetrabutylammonium sulphate ( to give final ratio aqueous/MeOH/CHCl3 3∶4∶8 ) . This was vortexed and centrifuged ( 1000g , 5 min ) to separate phases . The lower phase was removed into fresh tubes containing synthetic upper phase , mixed and centrifuged ( 1000 g , 5 min ) , then dried in vacuo . Dried lipids were deacylated by addition of monomethylamine reagent [41] , warmed ( 53°C , 5 min ) , vortexed and incubated ( 53°C , 25 min ) . Samples were subsequently cooled ( to RT ) and dried in vacuo . Deionized H2O and petroleum ether/n-butanol/ethyl formate [4∶20∶1 ( v/v ) ] were then added , then the mixture vortexed and centrifuged ( 1000 g , 5 min ) . The upper organic phase was then removed , the lower water-soluble phase mixed with 1 vol petroleum ether/n-butanol/ethyl formate [4∶20∶1 ( v/v ) ] , and the mixture vortexed and centrifuged as previously . The lower phase and interface was then dried in vacuo . Labelled lipid species were then separated for analysis by high-performance liquid chromatography ( HPLC ) . Dried lipids were resuspended in dH2O ( 2 ml; bath sonicated , and vortexed ) and filtered ( 0 . 45 µm ) . Samples were loaded onto a pre-equilibrated Whatman Partisphere 5SAX column ( 12 . 5 cm ) and developed at 1 mlmin−1 using a gradient of H2O ( A ) versus 1 . 25 M NaH2PO4 ( B ) [0 min , 0% B; 1 min , 1% B; 30 min , 6% B; 31 min , 15% B; 60 min , 25% B; 61 min , 33% B; 80 min , 60% B; 81 min , 100% B] . Fractions were collected every 30 s and scintillant added ( Packard ‘Ultima Gold’ ) . Lipid species were identified from the elution profile by comparison to labelled standards of known individual phosphoinositide species . PI ( 5 ) P levels were assayed as described [42] . Briefly , cultured cells were lysed into ice cold 5% ( v/v ) perchloric acid and incubated on ice prior to centrifugation ( 5000g , 10 min , 4°C ) . Lipids were extracted from the pellet using acidified chloroform/methanol and an aliquot removed for phospholipid quantification such that PI5P levels could be assessed relative to total cellular lipid [42] . The remainder was dried and resuspended in 1 vol CHCl3∶CH3OH∶ammonium formate [5∶10∶2 ( v/v/v ) , final concentration 50 mM ammonium formate ) . This was added to 0 . 1 vol neomycin beads [42] successively washed in 1 vol CHCl3∶CH3OH∶H2O [5∶10∶2 ( v/v/v ) ] , CHCl3:CH3OH:ammonium formate [5∶10∶2 ( v/v/v ) , final concentration 0 . 5 M formate] , and finally resuspended in 0 . 1 vol CHCl3:MeOH:ammonium formate [5∶10∶2 ( v/v/v ) , final concentration 50 mM formate] . Samples were incubated ( RT , 20 min ) , centrifuged ( 4000 g , 1min ) and the supernatant discarded . Beads were washed twice with 1 vol 50 mM ammonium formate buffer and lipids eluted twice with 2 M triethylbicarbonate ( TEAB ) to which phosphatidylserine was added as an inert carrier ( 14 . 25 µM ) . Beads were incubated with 0 . 5 vol 2 M TEAB and phosphatidylserine ( 1 h ) , centrifuged ( 4000 g , 2 min ) and supernatants collected . Beads were washed again with 0 . 2 vol 2 M TEAB and phosphatidylserine ( 1 h ) , and supernatants combined with the first . Tubes were dried in vacuo ( 60°C , overnight ) . Lipid phosphorylation with recombinant type IIα phosphatidylinositol kinase to determine PI ( 5 ) P mass was performed as previously [42] . Phosphorylation was initiated by addition of 2 µCi [γ-32P] ATP and 5 µM ATP to each sample , and PI ( 4 , 5 ) P2 extracted and separated by thin layer chromatography . PI ( 5 ) P levels were determined from a standard curved with reference to known concentrations of synthetic PI ( 5 ) P , and expressed relative to amount of total phospholipids present in each sample . | Critical to the onset of Salmonella infection is the ability of bacteria to force their own entry ( ‘invade’ ) into intestinal cells of their mammalian host from where they replicate , spread and cause damage . To achieve this invasion , Salmonella deliver a cocktail of proteins directly into host target cells . These proteins override host cell communications and remodel cell structure , tricking the normally dormant cells into engulfing the invaders . Although we are beginning to understand the functions of each delivered protein , little is known about how their activities are coordinated . Here we describe new techniques that systematically examine the interplay between the delivered bacterial proteins within the host cell . The results illuminate an unexpectedly complex network of interrelated relationships that must be precisely coordinated to promote bacterial invasion . The data provide new insights into how this important pathogen triggers invasion of host cells during infection . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"infectious",
"diseases/bacterial",
"infections",
"cell",
"biology/cytoskeleton",
"microbiology/cellular",
"microbiology",
"and",
"pathogenesis",
"cell",
"biology/cell",
"signaling"
] | 2008 | Deciphering Interplay between Salmonella Invasion Effectors |
The progressive loss of CD4+ T cell population is the hallmark of HIV-1 infection but the mechanism underlying the slow T cell decline remains unclear . Some recent studies suggested that pyroptosis , a form of programmed cell death triggered during abortive HIV infection , is associated with the release of inflammatory cytokines , which can attract more CD4+ T cells to be infected . In this paper , we developed mathematical models to study whether this mechanism can explain the time scale of CD4+ T cell decline during HIV infection . Simulations of the models showed that cytokine induced T cell movement can explain the very slow decline of CD4+ T cells within untreated patients . The long-term CD4+ T cell dynamics predicted by the models were shown to be consistent with available data from patients in Rio de Janeiro , Brazil . Highly active antiretroviral therapy has the potential to restore the CD4+ T cell population but CD4+ response depends on the effectiveness of the therapy , when the therapy is initiated , and whether there are drug sanctuary sites . The model also showed that chronic inflammation induced by pyroptosis may facilitate persistence of the HIV latent reservoir by promoting homeostatic proliferation of memory CD4+ cells . These results improve our understanding of the long-term T cell dynamics in HIV-1 infection , and support that new treatment strategies , such as the use of caspase-1 inhibitors that inhibit pyroptosis , may maintain the CD4+ T cell population and reduce the latent reservoir size .
HIV-1 progression to the AIDS stage within untreated patients usually takes many years . As HIV-1 infection progresses , the CD4+ T cell population declines slowly and the infected individual becomes progressively more susceptible to certain opportunistic infections and neoplasms . These are particularly common when CD4+ T cells reach a level below 200 cells/ul , which defines AIDS [1–7] . How HIV-1 infection induces progressive CD4+ T cell depletion is unclear [8] . One explanation is that the turnover rate of CD4+ T cells is significantly increased in HIV or simian immunodeficiency virus ( SIV ) infected subjects [9 , 10] . Therefore , massive activation of CD4+ T cells , which leads to more viral infection and cell death , might outrun the regeneration of T cells and cause progressive depletion . Another explanation is the failure of CD4+ memory T cell homeostasis during progressive HIV infection . This is possibly due to the destruction of the microenvironment of organs and tissues supporting T cell regeneration [3 , 11–14] . It remains unclear whether the impaired conformation of T cell regenerative tissues leads to the regeneration failure or it is merely a pathogenic reformation caused by HIV to promote viral replication . Mathematical models may shed light on how the complex interplay between the immune response and viral infection leads to overt immunodeficiency . Matrajt et al . used a model to analyze the simian-human immunodeficiency virus ( SHIV ) infection data in macaques [15] . They found that uninfected or bystander cell death accounts for the majority of CD4+ T cell death [15] . Mohri et al . studied the turnover of CD4+ T cells and found that T cell depletion is primarily induced by increased cellular destruction rather than decreased cellular production [16] . Kovacs et al . also showed that HIV does not impair CD4+ T cell production but increases T cell proliferation [17] . Using a model including the activation of resting CD4+ T cells , Ribeiro et al . found that HIV infection increases both the activation rate of resting CD4+ T cells and the rates of death and proliferation of activated CD4+ T cells [18] . Chan et al . showed that the rapid proliferation of CD4+ T cells provides more targets for infection and that preservation of CD4+ T cells in natural host monkeys is due to the limited CD4+ T cell proliferation [19] . Thus , CD4+ T cell depletion may be caused by the massive immune activation during chronic infection . However , a model by Yates et al . suggested that if immune activation drives T cell decline , then the predicted decline would be very fast , which is not consistent with the time scale of T cell depletion during chronic infection [20] . The above observations and analyses may explain T cell depletion but the long-term dynamics of CD4+ T cells have been neither simulated by models nor compared with patient data . In a recent study , Hernandez-Vargas and Middleton [21] developed a model including the infection of macrophages to explain the three stages of HIV infection . Fast infection of CD4+ T cells can explain the CD4+ T cell and viral load dynamics in the early stages , while slow infection of macrophages may explain the dynamics in the advanced stages of infection . Whether macrophages form a long-term reservoir causing T cell depletion and viral explosion in the later stages of infection needs further experimental investigation . Different from apoptosis , a programmed process that results in non-inflammatory cell death , pyroptosis is a form of programmed cell death associated with antimicrobial responses during inflammation [22] . During HIV infection , Doitsh et al . [23 , 24] found that when virus enters a CD4+ T cell that is non-permissive to viral infection , the caspase-1 pathway is triggered to induce pyroptosis , which can secrete inflammatory cytokines such as IL-1β . These cytokines establish a chronic inflammation state and attract more CD4+ T cells to the inflamed sites , resulting in more infection and cell death . Thus , pyroptosis generates a vicious cycle in which dying CD4+ T cells secrete inflammatory signals that attract more CD4+ T cells to be infected and die [23] . These findings suggest that HIV-1 may use the intrinsic feature of the immune system to seek targets of infection , establish productive viral replication , and meanwhile destroy the CD4+ T cell population . Here we developed mathematical models incorporating the effect of pyroptosis to study whether it can explain the very slow T cell depletion during HIV-1 infection . Using the models we explored if highly active antiretroviral therapy ( HAART ) can preserve the CD4+ T cell population . We studied the effect of CD4+ T cell proliferation and CD8+ T cell response on CD4+ decline . We also compared our modeling prediction with clinical data obtained from patients in Rio de Janeiro , Brazil [25–28] . At last , we probed the possible contribution of chronic inflammation associated with pyroptosis to the HIV latent reservoir persistence .
The patient data were obtained from seroconverters in 3 cohorts [25–28] . One cohort consists of high-risk , HIV-seronegative homosexual and bisexual men who did not report injection drug use , were enrolled between July 1995 and June 1998 and seroconverted during follow-up [26] . The other cohorts consist of seroconverters from high-risk HIV-seronegative homosexual and bisexual men patients who were enrolled from December 1998 to May 2001 in a study designed to evaluate the behavior impact of post-exposure prophylaxis [27] , and participants from the control arm of SPARTAC , a randomized trial designed to evaluate the impact of short term antiretroviral therapy on the course of primary HIV infection [28] . The median of the CD4+ T cell data was derived from these cohort studies . The median of the Current Study Multicenter AIDS Cohort Study ( MACS ) was obtained from the study [29] . These patient data and medians were compared with modeling prediction . Inflammatory cytokines released by abortively HIV-infected cells can attract more CD4+ T cells to be infected . In the following one-compartment model , to minimize the number of variables and parameters we described the effect of pyroptosis by use of an enhanced viral infection rate because of increased availability of CD4+ T cells attracted by cytokines to the inflamed sites . The variable T represents the population of uninfected CD4+ T cells . They are generated at the rate λ . Proliferation of target cells will be considered later . The infection rate is modeled by a mass action term kVT , which is enhanced by the inflammatory cytokine ( C ) with a factor γi . Uninfected T cells die at a per capita rate d1 . T* is the population of productively infected T cells and their death rate is d2 . A fraction ( f ) of new infection is assumed to be abortively infected . The death rate of abortively infected T cells ( M* ) is d3 . Virus ( V ) is generated by productively infected T cells with a viral production rate pv and is cleared at a rate d4 . Inflammatory cytokines are released with a burst size ( Nc ) when an abortively infected cell dies . Thus , Ncd3 represents the generation rate of cytokines per abortively infected cell . The decay rate of cytokines is assumed to be d5 . The schematic diagram of this model is shown in Fig 1 . Parameters and values are listed in Table 1 . In the above one-compartment model , we described the consequence of pyroptosis but did not explicitly model the cytokine-induced attraction of CD4+ T cells from elsewhere to the place where abortive infection occurs . Below we developed another model with two compartments to include cytokine-induced T cell movement explicitly . The model is more complicated and contains more parameters . In the model there are two compartments: one represents the blood ( T1 ) and the other represents human lymphoid tissues ( T2 ) such as lymph nodes in which abortive infection takes place on a large scale [23] . CD4+ T cells in compartment I ( or II ) can transport to compartment II ( or I ) at a rate σ1 ( or σ2 ) . In blood , cytokines released during abortive infection cannot accumulate as in lymphoid tissues . They cannot attract other immune cells to fight the infection and contribute to inflammation . Thus , pyroptosis is assumed to take place only in lymphoid tissues ( compartment II ) , as observed in ref . [23] . The transportation rate σ1 from the blood to tissues is assumed to be enhanced by a factor ( 1+γrC ) due to inflammatory cytokines ( C ) released during pyroptosis in compartment II . Viruses ( V1 and V2 ) can also transport between two compartments with the rates D2 ( V1-V2 ) and D1 ( V2-V1 ) , which depend on the difference of viral load in the two compartments . Because the dynamics of the virus are much faster than those of infected cells , it is reasonable to assume that they are proportional to each other . Thus , we only included the transportation of virus between compartments . In the Supporting Information ( S1 Text and S7 Fig ) , we added the transportation of infected cells to the model and found that the model prediction is similar to the case without infected cell transportation . All the other variables and parameters ( summarized in Table 1 ) can be defined similarly as those in the one-compartment model ( Fig 1 ) . The schematic diagram of the two-compartment model is shown in Fig 2 . For model simulation , we fixed most of parameters based on existing experimental data and our previous modeling studies [30–33] . Because the CD4+ T cell level within an uninfected individual ranges normally from 500 cells/μl to 1500 cells/μl , we changed the unit to cells/ml and assumed CD4+ T cells to be 106 cells/ml before infection [34] . The death rate ( d1 ) of uninfected CD4+ T cells is assumed to be 0 . 01 day-1 [35] . Thus , from the steady state of target cells before infection , we obtained that the generation rate ( λ ) of target cells is 106 ( 0 . 01 ) = 104 cells ml-1 day-1 . The viral infection rate k is assumed to be 2 . 4×10−8 ml virion-1 day-1 [30] . The death rate of infected T cells is d2 = 1 day-1 [36] . We chose the parameter γi to be 2×10−4 ml molecule-1 . The viral production rate of productively infected T cells in the one-compartment model is chosen to be 2 . 5×104 virions cell-1 day-1 [37] . As described by Doitsh et al . [23 , 24] , abortive infection accounts for 95% of the total infection . Thus , we chose f to be 0 . 95 . Because abortive infection mainly takes place in non-permissive quiescent T cells , we chose their death rate ( d3 ) to be 0 . 001 day-1 [31 , 32] . The burst size of cytokines is fixed to Nc = 15 molecules . The half-life of IL-1β is about 2 . 5 hours [38] . Thus , we chose the decay rate of cytokines to be d5 = 6 . 6 day-1 . We also performed sensitivity tests of the modeling prediction on a number of parameters . We fit both the one-compartment and two-compartment models to subjects with more than 10 data points [25–29] . The root mean square ( RMS ) between model prediction and patient data is minimized for each patient . RMS is calculated using the following formula RMS=Σi=1n ( T ( ti ) +T* ( ti ) −T^ ( ti ) ) 2n where T ( ti ) +T* ( ti ) represents the CD4+ T cell population level in blood at time ti predicted by the model , T^ ( ti ) is the corresponding patient data at ti . We used T1 ( ti ) +T1* ( ti ) in the fitting for the two-compartment model . Parameter estimates are based on the best fit that achieves the minimum RMS . Data fitting is performed using the R programming language . In order to statistically compare the best fits of using the two models , we calculated the Akaike information criterion ( AIC ) . The model with a lower AIC value fits the data better from a statistical viewpoint . The AIC is calculated using the following formula AIC=nln ( RSSn ) +2m RSS=∑i=1n ( T ( ti ) +T* ( ti ) −T^ ( ti ) ) 2 where n is the number of observations ( i . e . number of data points ) and m is the number of fitted parameters . RSS is the residual sum of squares . T ( ti ) , T* ( ti ) and T^ ( ti ) are the same as those defined in the calculation of RMS . We obtained the 95% confidence intervals for fitted parameters using a bootstrap method [39] , where the residuals to the best fit were re-sampled 200 times .
Using the parameter values listed in Methods and initial values V ( 0 ) = 1×10−3 RNA copies/ml , T ( 0 ) = 103 cells/μl , T* ( 0 ) = 0 , M* ( 0 ) = 0 , and C ( 0 ) = 0 in the one-compartment model , we showed that the population of CD4+ count declines from 103 cells/μl to about 200 cells/μl around the 6th year after infection ( Fig 3A ) . This is consistent with the slow time scale of T cell decline during HIV infection . The entire T cell depletion course consists of two major phases . The first massive depletion phase is rapid , followed by a slower chronic depletion phase ( Fig 3A ) . The first-phase T cell decline is due to the substantial viral infection during the early stage . If there is no infection ( k = 0 ) , then the T cell level would stabilize at the initial level ( Fig 3A ) . The slow second-phase T cell decline is due to pyroptosis enhanced viral infection . Without the effect of inflammatory cytokines released during pyroptosis ( i . e . γi = 0 or no inflammation in Fig 3A ) , a balance between T cell generation and viral infection is reached and the T cell population is maintained at a steady state level . This agrees with the prediction of most viral dynamics models without treatment . Because of pyroptosis , cytokine-enhanced viral infection breaks the balance between cellular production and viral infection , which makes the T cell level decline at a very low rate and approach the immune-deficient level after several years ( Fig 3A ) . The viral load change was plotted in Fig 3B . Without the effect of inflammatory cytokines , the viral load reaches a steady state level . When there is cytokine enhanced viral infection , viral load increases very slowly during the phase of chronic infection ( Fig 3B ) . Using a constant λ is a simple way to approximate the generation of target cells . We included the proliferation of target cells in the model ( S1 Text ) . Simulation with different proliferation rates is shown in S1 Fig . As the proliferation rate increases , the decline of CD4+ T cells becomes faster . This is because more target cells lead to more abortive infection , which releases more cytokines attracting more CD4+ T cells to be infected and die . This prediction is consistent with the observation that the level of T cell proliferation in non-pathogenic infection ( e . g . SIV infection in natural host monkeys such as sooty mangabeys or mandrills that do not develop AIDS-like diseases ) was much lower than in pathogenic infection , e . g . , SIV in rhesus macaques [40 , 41] . This provides an additional support to the view that an attenuated rather than effective adaptive immune response preserves immune function in natural host monkeys [42] . We performed sensitivity analysis of the CD4+ T cell decline for a number of parameters . Fig 4 shows that the sensitivity tests on parameters k , λ , pv and γi . S2–S5 Figs show the tests on parameters Nc , d3 , d5 , and f , respectively . We found that the model is robust in generating the slow decline of CD4+ T cells , although the model prediction is more sensitive to three parameters k , pv and f ( see Figs 4A , 4C and S5 ) . In the above simulation , we assumed that the viral infection enhancement parameter γi is a constant . When the concentration of inflammatory cytokines is low , they may not be able to trigger the attraction of CD4+ T cells from elsewhere . Thus , we simulated a scenario in which enhanced viral infection is triggered only when the level of cytokines is above a threshold value . We chose γi to be the following step function . It is zero if the level of cytokines is below a certain level . The threshold value was chosen to be 2000 or 4000 molecules/ml in Fig 5A . CD4+ T cells do not decline until the level of cytokines reaches the corresponding threshold ( Fig 5B ) . A more realistic scenario is that γi increases gradually when the concentration of cytokines is above the threshold . We chose γi ( C ) to be the following exponential function . The hill coefficient ρ determines how fast γi ( C ) increases from 0 to its maximum value γi . Both ρ and γi were fixed to 2×10−4 ml molecule-1 . With a non-constant parameter γi ( C ) , we found that CD4+ T cells also undergo a slow decline to below 200 cells/μl ( Fig 5B and 5C ) . Using an exponential function for γi ( C ) , the decline of CD4+ T cells is smoother than the case using a step function . Using the one-compartment model , we studied if HAART can rescue the CD4+ T cell population . During HAART we assumed that the viral infection rate k is reduced by a factor ( 1-ε ) , where ε is the overall drug efficacy of the treatment [32] . The simulation shows that if the treatment effectiveness is very high , then CD4+ count can rebound to its pre-infection level ( Fig 6A ) no matter when HAART is initiated . For lower treatment effectiveness ( e . g . ε = 0 . 6 in Fig 6B ) , the patient needs a relatively long time to restore the CD4+ T cell population . The later HAART starts , the longer it takes for CD4+ T cell restoration ( Fig 6B ) . When the treatment effectiveness is further lower , CD4+ T cell depletion could not be prevented . These results suggest that HAART has the potential to rescue CD4+ T cell population , but CD4+ response depends on the effectiveness of the therapy and when the therapy is initiated . We included CD8+ T cells in the one-compartment model to study the interaction between CD4+ T cell decline and CD8+ T cell response . CD8+ T cells ( E ) are assumed to kill infected T cells at a rate αET* . The activation rate of CD8+ T cells depends on the level of infected cells with a half-maximal saturation constant θ . pE is the maximum activation rate . CD4+ T cells play an important role in activating the adaptive immune response . We used another saturation function T/ ( T+η ) to account for this influence . The T* and E equations are given below . The simulation of the model with CD8+ T cell response is shown in Fig 7 . Parameter values are listed in Table 1 . For comparison , we plotted the predicted T cell dynamics with and without the influence of CD4+ T cells . In column A of Fig 7 , we performed the simulation without T/ ( T+η ) . CD4+ T cells decline slowly and CD8+ T cells reach a steady state level . Column B shows the simulation with the term T/ ( T+η ) . CD8+ T cell response becomes weaker than in column A because of the slow depletion of CD4+ T cells . CD4+ T cells decline faster because of the incapability of CD8+ T cells to control viral infection . Inflammatory signals released during pyroptosis induce the movement of CD4+ T cells from circulation in blood to inflamed lymph nodes [43–46] . We developed a more comprehensive model by including two cell compartments ( Fig 2 ) . One is the blood compartment and the other is the compartment of lymphoid tissues where pyroptosis takes place . Simulation of the two-compartment model shows that CD4+ count in blood declines from 103 cells/ul to 200 cells/ul over a long time period ( Fig 8A ) . The viral load change in blood is also similar to that shown in Fig 3B except that T cell and viral load dynamics generated by the two-compartment model have less oscillation than by the one-compartment model . Using the two-compartment model we also tested if HAART can rescue CD4+ T cell population . We assumed that the drug efficacies of HAART within blood and lymph node are different ( i . e . , the viral infection rate k in compartment I is reduced by 1-ε1 and k in compartment II is reduced by 1-ε2 ) . We found that if the drug efficacies in both compartments are high , then CD4+ T cell depletion can be prevented ( Fig 8B ) . The time for CD4+ restoration also depends on when HAART is initiated . However , if the drug efficacy in compartment II is relatively low ( e . g . ε2 = 0 . 4 ) compared with the high efficacy in compartment I ( e . g . ε1 = 0 . 9 ) , then CD4+ T cells decline even when HAART is initiated at the beginning of viral infection ( Fig 8C ) . In the simulation , CD4+ T cells stabilize at 230 cells/ul after more than 30 years ( Fig 8C ) . This result suggests that even if some lymphoid tissues might be difficult for drug's penetration ( i . e . drug sanctuary sites ) , CD4+ T cells can be maintained at a higher level in treated patients than in untreated patients . This may explain the increased life expectancies in HIV patients treated with combination therapy [47–51] . However , because of the CD4+ cell decline ( Fig 8C ) , life expectancy should be lower in patients with lower baseline CD4+ cell counts than in those with higher baseline counts . This is consistent with the reported life expectancy of individuals on combination therapy in a collaborative analysis of 14 cohort studies [47] . We compared modeling predictions with the CD4+ T cell data shown in [25–29] . Using the one-compartment model , we fit parameters k , γi , λ , pv and fix the other parameters for each patient . We also fit the model to the median data calculated from all the patients in the two cohort study [25] and the median data of the Current Study Multicenter AIDS Cohort Study ( MACS ) [29] . Using the two-compartment model , we fit parameters k , γr , λ1 , pv1 to the same patient and median data . Figs 9 and 10 show that both models provide a good fit to the long-term CD4+ T cell data in untreated HIV-1 patients . The fit to the median data is better than the fit to individual patients based on the calculated error between modeling prediction and data . These data fits suggest that pyroptosis induced CD4+ T cell movement during abortive infection can explain the progressive CD4+ T cell depletion observed in untreated HIV-1 patients . Parameter estimates and their 95% confidence intervals based on the fits to the one-compartment and two-compartment models are listed in Tables 2 and 3 , respectively . The estimate of the viral production rate pv in the one-compartment model is higher than the viral production rate pv1 in blood of the two-compartment model ( pv2 = 2000 virions per cell per day is fixed during fitting ) . This is because in the one-compartment model 95% of infection is assumed to be abortive and only 5% of infection produces virus . Thus , a higher value of viral production rate is needed to generate viral load with reasonable magnitude . In the second-compartment model , although only 5% of infection produces viruses in lymphoid tissues , the target cell level is much higher in lymphoid tissues than in blood ( i . e . λ2 >> λ1 ) . Thus , the viral production rates in the two compartments are on the same order of magnitude . The Akaike information criterion ( AIC ) value is calculated to compare data fitting using the two models ( Tables 2 and 3 ) . We found that for patients 11 , 38 , 44 , and median of patient data , the AIC value of using the second model is less than that of using the first model . This suggests that the two-compartment model provides a better fit to the data for these patients from a statistical viewpoint . IL-7 plays an important role in latently infected CD4+ T cell proliferation [52] . It has been observed to be over expressed in inflamed tissues [53 , 54] . Inflammatory cytokines released during cell death by pyroptosis may promote the establishment and persistence of the latent reservoir in HIV patients . Here we included the population of latently infected CD4+ T cell ( L ) into the one-compartment model . Latently infected CD4+ cells are produced with a fraction μ during HIV-1 infection . They can also be maintained by proliferation which is assumed to rely on the cytokine level ( see the term 1+φC in the following equation where φ is fixed to 10−2 ml molecule-1 ) . We chose the base proliferation rate pL to be 0 . 001 day-1 [32] , which represents a limited proliferation capacity in the absence of inflammatory cytokines . The carrying capacity of latently infected cells ( Lmax ) is fixed at 100 cells/ml [32] . The other parameter values are listed in Table 1 . The equations of L and T* are given below and the other equations are the same as those in the one-compartment model . If there is no chronic inflammation ( i . e . φ = 0 in the L equation ) , then latently infected cells undergo a slow decline ( Fig 11A ) . However , if the proliferation is enhanced by cytokines released during cell death by pyroptosis , then the latent reservoir can be maintained at a higher level ( Fig 11B ) . This result suggests that inflammatory cytokines generated during abortive infection might contribute to the establishment of the latent reservoir and the maintenance of its size . We also performed sensitivity test of latently infected cells on the parameter φ , the effectiveness of cytokines promoting latently infected cell proliferation . The modeling prediction is robust to this parameter ( S6 Fig ) . Latently infected cells can be activated by relevant antigens and become productively infected cells . In S1 Text , we included the activation of latently infected cells in the one-compartment model . Simulation with different values of the activation rate is shown in S8 Fig . As the activation rate increases , the size of the latent reservoir decreases .
The mechanisms underlying the slow time scale of CD4+ T cell decline in untreated HIV-1 patients remain unclear . HIV-mediated cell death can contribute to the loss of CD4+ T cells , but quantitative image analysis suggested that infection-induced cell death could be compensated by upregulated T cell division [55 , 56] . Some studies suggested that the destruction of bystander non-infected cells may account for the CD4+ T cell decline during disease progression [57–60] . Immune activation might be the reason of bystander cellular demise [60] . It drives uninfected CD4+ T cells into several rounds of division and cells are susceptible to activation-induced death [61 , 62] . However , a mathematical model showed that the decline of CD4+ T cells would be very rapid if immune activation drives T cell depletion [20] . Another possible reason of T cell decline might be the regeneration failure of CD4+ T cells during disease progression [3 , 11–14] . A recent study found that about 95% of CD4+ T cells within lymph nodes die from pyroptosis and release inflammatory signals that attract more CD4+ T cells from elsewhere to be infected [23] . HIV-1 may use this vicious infection cycle to promote disease progression and chronic T cell depletion . In this paper , we developed mathematical models to explore whether cell death induced by pyroptosis can explain the slow time scale of CD4+ T cell decline in untreated HIV patients . In the first model , we assumed that increased availability of target cells due to attraction by inflammatory cytokines facilitates viral infection , which drains the CD4+ T cell population slowly during chronic infection . In the second model , we explicitly included the movement of CD4+ T cells from blood to lymphoid tissues where pyroptosis occurs . Both models generate a very slow decline of CD4+ T cells in plasma ( Figs 3 and 8 ) , and agree with the long-term CD4+ T cell data from untreated HIV patients in several cohorts in Brazil ( Figs 9 and 10 ) . We found that the entire CD4+ T cell decline consists of two major phases ( Fig 3 ) . The first-phase decline is very rapid . This decline is due to the enormous virus infection and virus-induced cell death during primary infection . Following the first phase , CD4+ T cells partially recover because of cell regeneration and viral control by immune responses . However , a balance cannot be established between cell generation and viral infection . Chronic inflammatory cytokines released during pyroptosis can attract CD4+ T cells from other places to inflamed lymphoid tissues . These cells are infected and die , resulting in a slow decline of CD4+ T cells in plasma . These results suggest that HIV-mediated cell death causes the dramatic decline of CD4+ T cells during primary infection and that persistent chronic inflammation acts like an erosive force which gradually drains the CD4+ T cell population in plasma during chronic infection . HAART was shown to have the potential to restore the CD4+ T cell population ( Figs 6 and 8 ) , which agrees with the robust and sustained CD4 recovery among patients remaining on therapy [63] and a normal life expectancy in patients with a good CD4 response and undetectable viral load [50] . However , CD4 response depends on the effectiveness of the therapy , when the therapy is initiated , and whether there exist drug sanctuary sites ( Figs 6 and 8 ) . This may explain the considerable variability in the increase of life expectancy in patients treated with combination therapy between 1996 and 2005 [47] . Our model has limitations . First , it does not account for the spatial effect of CD4+ T cells . Although we used a two-compartment model to describe the transportation of cells and virus between blood and lymphoid tissues , release of cytokines during cell death by pyroptosis and attraction of CD4+ T cells are mainly constrained to occur locally . Ordinary differential equation models could not capture these features . It would be valuable to develop spatial models that can describe the vicious cycle within lymphoid tissues . Spatial models require precise description and parameterization of diffusion of cytokines and attraction of CD4+ T cells , and are also computationally demanding in studying T cell dynamics within blood and different lymphoid tissues . The second limitation of our model is that we did not consider a detailed inflammatory signal transduction cascade between T cells and relevant tissues . Recruitment of T cells to the inflamed tissue goes through several steps of immunological reaction . Upon secretion of IL-1β , expression of adhesion molecules such as E/P-selectin and ICAM-1 on the vascular endothelium is upregulated [64] . Binding to these molecules facilitates T cell's attachment to vascular endothelium . After attachment T cells undergo conformational changes and penetrate into the inflamed tissue [65 , 66] . In our models , we used a very simple factor multiplied by the concentration of cytokines to describe the effect of inflammatory cytokines . A more comprehensive model requires a detailed description of intracellular processes underlying the inflammatory signal cascade and related data for model verification . The third limitation is that our model cannot generate viral load explosion in the later stages of HIV infection . Assuming that all parameters are constant and that only one cell population produces virus , our model cannot describe viral explosion . However , as CD4+ T cells drop to very low levels , the immune system cannot kill infected cells or neutralize virus effectively . This leads to a reduction in the death rate of infected cells or viral clearance rate , and may explain the viral explosion . Infection of other cell populations such as macrophages ( as suggested by Hernandez-Vargas and Middleton in ref . [21] ) or other viral reservoirs may also explain the dramatic viral load increase during the AIDS stage . Our simulation shows that the latent reservoir may be maintained by chronic inflammation . How inflammation promotes the latent reservoir persistence is not fully understood . Some results suggested that caspase-1 can promote cellular survival . For example , epithelial cells activate caspase-1 to enhance membrane repair in response to the pore-forming toxins to prevent proteolysis [67] . Whether latently infected T cells can use this caspase-1 pathway to promote their survival remains unknown . Another possibility is through the dysregulated action of IL-7 or IL-15 that can stimulate homeostatic proliferation of latently infected cells . Stromal cells are located in secondary lymph organs such as lymph node trabeculae , lymph vessels , and conduits [68] . IL-7 is observed to be significantly expressed by stromal cells within inflamed lymph nodes [69] . It would be valuable to explore whether HIV-1 can use the caspase-1 pathway to persist in latent cells and whether IL-7 production can be inhibited in the inflamed microenvironment . The results suggest that cell death by pyroptosis plays an important role in driving slow CD4+ T cell depletion . If pyroptosis can be inhibited , then CD4+ T cells might be maintained . VX-765 is a caspase-1 inhibitor [70–73] used to treat chronic epilepsy and psoriasis . It was found to be safe and well tolerated in humans in a phase IIa trial of epilepsy [74] . Doitsh et al . showed that VX-765 can inhibit secretion of IL-1β and also block cleavage of caspase-1 in HIV-infected tonsillar and splenic lymphoid tissues [23] . However , the active form of VX-765 cannot effectively inhibit cell death , which may be due to reduced cellular permeability [70] . It remains unclear whether the pro-drug VX-765 can efficiently block cell death in vivo . We showed that if antiretroviral drugs cannot effectively block viral replication in lymphoid tissues , then HIV-1 can still establish chronic inflammation in these sites . This is consistent with the observation of persistent inflammation in patients under long-term antiretroviral treatment [75 , 76] . When drug sanctuary sites exist , CD4+ T cells undergo a very slow depletion or stabilize at a low level ( Fig 8 ) . In this case , the immune system would be vulnerable to various opportunistic infections and neoplasms . This may explain the shorter extension of life expectancy in treated patients who had a low CD4+ cell nadir [47–49 , 51] . If antiretroviral drugs and caspase-1 inhibitors can be effectively delivered to human lymphoid tissues via some transporters [77 , 78] , then CD4+ T cell depletion might be prevented and life expectancy of treated patients might be further extended . | The CD4+ T cell population within HIV-infected individuals declines slowly as disease progresses . When CD4+ cells drop to below 200 cells/ul , the infection is usually considered to enter the late stage , i . e . , acquired immune deficiency syndrome ( AIDS ) . CD4+ T cell depletion can take many years but the biological events underlying such slow decline are not well understood . Some studies showed that the majority of infected T cells in lymph nodes die by pyroptosis , a form of programmed cell death , which can release inflammatory signals attracting more CD4+ T cells to be infected . We developed mathematical models to describe this process and explored whether they can generate the long-term CD4+ T cell decline . We showed that pyroptosis induced cell movement can explain the slow time scale of CD4+ T cell depletion and that pyroptosis may also contribute to the persistence of latently infected cells , which represent a major obstacle to HIV eradication . The modeling prediction agrees with patient data in Rio de Janeiro , Brazil . These results suggest that a combination of current treatment regimens and caspase-1 inhibitor that can inhibit pyroptosis might provide a new way to maintain the CD4+ T cell population and eradicate the HIV latent reservoir . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [] | 2015 | Modeling the Slow CD4+ T Cell Decline in HIV-Infected Individuals |
The nuo-6 and isp-1 genes of C . elegans encode , respectively , subunits of complex I and III of the mitochondrial respiratory chain . Partial loss-of-function mutations in these genes decrease electron transport and greatly increase the longevity of C . elegans by a mechanism that is distinct from that induced by reducing their level of expression by RNAi . Electron transport is a major source of the superoxide anion ( O⋅– ) , which in turn generates several types of toxic reactive oxygen species ( ROS ) , and aging is accompanied by increased oxidative stress , which is an imbalance between the generation and detoxification of ROS . These observations have suggested that the longevity of such mitochondrial mutants might result from a reduction in ROS generation , which would be consistent with the mitochondrial oxidative stress theory of aging . It is difficult to measure ROS directly in living animals , and this has held back progress in determining their function in aging . Here we have adapted a technique of flow cytometry to directly measure ROS levels in isolated mitochondria to show that the generation of superoxide is elevated in the nuo-6 and isp-1 mitochondrial mutants , although overall ROS levels are not , and oxidative stress is low . Furthermore , we show that this elevation is necessary and sufficient to increase longevity , as it is abolished by the antioxidants NAC and vitamin C , and phenocopied by mild treatment with the prooxidant paraquat . Furthermore , the absence of effect of NAC and the additivity of the effect of paraquat on a variety of long- and short-lived mutants suggest that the pathway triggered by mitochondrial superoxide is distinct from previously studied mechanisms , including insulin signaling , dietary restriction , ubiquinone deficiency , the hypoxic response , and hormesis . These findings are not consistent with the mitochondrial oxidative stress theory of aging . Instead they show that increased superoxide generation acts as a signal in young mutant animals to trigger changes of gene expression that prevent or attenuate the effects of subsequent aging . We propose that superoxide is generated as a protective signal in response to molecular damage sustained during wild-type aging as well . This model provides a new explanation for the well-documented correlation between ROS and the aged phenotype as a gradual increase of molecular damage during aging would trigger a gradually stronger ROS response .
Mitochondrial function has been linked to the aging process in a number of ways [1] . In particular , mitochondria are crucial in energy metabolism and as such have been implicated in the aging process by one of the very first theories of aging [2] , the rate-of-living theory of aging [3] , which suggested that the rate of aging is proportional to the rate of energy metabolism ( reviewed in [4] ) . Mitochondrial function in animals is also known to decline with age [5] , [6] , which , together with the finding that mitochondria are an important source of toxic reactive oxygen species ( ROS ) , has led to the oxidative stress ( or free radical ) theory of aging [7] , [8] . Two types of mutations that affect mitochondrial function have been found to affect the rate of aging in C . elegans , mutations that shorten lifespan , such as mev-1 [9] and gas-1 [10] , and mutations that lengthen lifespan , such as clk-1 [11] , isp-1 [12] , lrs-2 [13] , and nuo-6 [14] . lrs-2 encodes a mitochondrial leucyl-tRNA-synthetase , and its effect on the function of mitochondrial electron transport is likely relatively indirect , via partial impairment of mitochondrial translation . However , clk-1 encodes an enzyme necessary for the biosynthesis of ubiquinone , a lipid antioxidant and an electron transporter of the respiratory chain [15] , and mev-1 , gas-1 , isp-1 , and nuo-6 all encode subunits of mitochondrial respiratory complexes . On the strength of the oxidative stress theory of aging it has been suggested , and supported by a number of observations ( reviewed in [16] , [17] ) , that the mev-1 and gas-1 mutations reduce lifespan by increasing mitochondrial oxidative stress , and clk-1 , isp-1 , and nuo-6 increase lifespan by reducing it . In addition to genomic mutations that affect mitochondrial proteins , it has been found that knockdown by RNA interference of C . elegans genes that encode subunits of mitochondrial complexes , including isp-1 and nuo-6 , also prolongs lifespan [13] , [18] , [19] . Although the effect of RNAi on ETC subunits , which is conserved in Drosophila [20] , was initially believed to be similar to that of the mutations [21] , [22] , [23] , it was recently found that it is in fact distinct and separable [14] . A recent study analyzed patterns of gene expression in isp-1 mutants together with those in clk-1 and cyc-1 ( RNAi ) [23] and suggested that the overlap between these patterns could define the biochemical processes that underlie the effect of all interventions that impact mitochondria . However , our recent findings that isp-1 ( qm150 ) and isp-1 ( RNAi ) trigger fully separable mechanisms suggests that the overlapping gene expression changes identified by Cristina et al . [23] might not be sufficient to prolong lifespan . Rather some of the gene expression changes that are specific to each type of intervention are necessary for their effect on lifespan and can act additively . isp-1 mutants show a trend toward low levels of oxidative damage to proteins , increased expression of the cytoplasmic Cu/Zn superoxide dismutase ( SOD-1 ) and of the mitochondrial Mn superoxide dismutase ( SOD-2 ) [24] , and increased resistance to acute treatment with the prooxidant paraquat [14] . However , although knocking down the genes encoding the major superoxide dismutase by RNAi results in normal or elevated levels of oxidative damage , it had no effect on the lifespan of the mutants [24] , suggesting that the reduced oxidative damage found in isp-1 mutants is not responsible for their longevity . Furthermore , the notion that mitochondrial oxidative stress could be the cause of aging has recently been challenged by a number of studies in C . elegans [24] , [25] , [26] , [27] , [28] , in Drosophila [29] , and in mice ( reviewed in [30] ) . ROS are not just toxic metabolites that lead to oxidative stress but are also signaling molecules that are believed to be involved in a mitochondria-to-nucleus signaling pathway that could impact aging [1] , [31] , [32] , [33] . Interfering with mitochondrial function has the potential to alter the rate and/or the pattern of production of ROS by mitochondria , including in counter-intuitive ways . For example , reducing oxygen concentration increases ROS production by mitochondrial complex III in vertebrate cells [34] , [35] , and the knockout of sod-2 in C . elegans can lead to normal [25] or increased lifespan in spite of increased oxidative damage [26] . Here we examined ROS production by mitochondrial mutants and found that isp-1 and nuo-6 mutants have increased generation of the superoxide anion but not increased levels of other ROS and that this increase is necessary and sufficient for longevity , suggesting that superoxide triggers mechanisms that slow down aging , presumably at the level of gene expression .
To measure changes in mitochondrial ROS generation that could affect signaling , it is not adequate to measure the level of ROS damage , as a change in ROS damage levels can be brought about by changes in detoxification of ROS , in protein turnover , or in damage repair . However , it is notoriously difficult to directly visualize or measure ROS generation and ROS levels in intact organisms including in living worms . To overcome this difficulty we have adapted a technique originally developed for vertebrates that uses flow cytometry to sort isolated intact mitochondria and measure ROS levels with indicator dyes ( Figure S1 ) [37] . Mitochondria were extracted from worms by standard techniques and loaded with either one of two fluorescent indicator dyes , H2DCFDA , a dye that is sensitive to a variety of ROS but rather insensitive to superoxide [38] , [39] , and MitoSox , a dye that is exclusively sensitive to superoxide [40] . The prooxidant paraquat ( PQ ) induces mitochondrial superoxide generation [41] , and the antioxidant N-acetyl-cysteine ( NAC ) has an antioxidant effect on all types of ROS [42] , [43] . As expected , when purified mitochondria were treated with PQ , the fluorescence of both H2DCFDA and MitoSox increased , and the fluorescence of both decreased when treated with NAC ( Figures 1A , 1B , and S1B ) . One limitation of this technique is the need for a rather large amount of mitochondria . For example , a sufficient amount of worms is not readily obtained from worms treated by RNAi , and we have therefore focused on long-lived mutants only . We used the cytometry technique to determine the generation of mitochondrial superoxide and of overall mitochondrial ROS in a number of long-lived mutants . Both isp-1 and nuo-6 mutations did not affect H2DCFDA fluorescence ( overall ROS ) significantly , but both showed elevated MitoSox fluorescence ( superoxide ) ( Figure 1C and 1D ) . Mutants of four other genes ( clk-1 ( qm30 ) , eat-2 ( ad1116 ) , daf-2 ( e1370 ) , and sod-2 ( ok1030 ) ) were also tested ( Figure 1E and 1F ) . clk-1 mutants showed an elevation of overall ROS-associated fluorescence but not of superoxide-associated fluorescence . daf-2 mutants were most similar to the mitochondrial respiratory chain mutants with an elevation of superoxide-associated fluorescence but no significant elevation in overall ROS-associated fluorescence . Finally , eat-2 and sod-2 mutants showed no significant elevation in either signal but only a trend for low overall ROS in the case of eat-2 mutants and a trend for increased superoxide in the case of sod-2 mutants . The elevated MitoSox signal in isp-1 , nuo-6 , and daf-2 corresponds mostly to increased superoxide generation , as all three mutants are known for elevated levels of the mitochondrial SOD-2 and SOD-3 [12] , [14] , [24] , [44] , whose activity would prevent the accumulation of superoxide . Elevated superoxide detoxification , however , should not prevent measuring increased superoxide generation as superoxide is generated at prosthetic electron carriers such as ubiquinone in complex III [45] , [46] and FMN in complex I [47] , [48] , which are at least partially buried in the complexes . Thus a small molecular weight dye that has access to these sites can trap the superoxide before it has the opportunity to diffuse toward the SOD-2 and SOD-3 proteins . There is no increase in the H2DCFDA signal in these mutants likely because this dye is not particularly sensitive to superoxide [49] . It appears therefore that in the presence of efficient detoxification the level of overall ROS is not significantly increased by the increased superoxide generation that we observe . This is consistent with the finding that these mutants do not have increased oxidative damage [14] , [24] . sod-2 deletion mutants do not show a significant increase in the MitoSox signal ( Figure 1E and 1F ) , indicating that decreased detoxification does not lead to an easily measurable increase in this signal in purified mitochondria . The signal from H2DCFDA , a dye which has very broad sensitivity but is not very sensitive to superoxide [49] , is also unchanged , suggesting that , at least in isolated worm mitochondria , electron transport is not the main source of the type of ROS to which H2DCFDA dye is significantly sensitive . The level of superoxide generation in these mutants might also be kept moderately low because of their reduced electron transport [26] , although low electron transport could in principle also result in elevated superoxide as we have observed in isp-1 and nuo-6 mutants . clk-1 mutants have only a small deficit in electron transport [24] , [50] , [51] , in spite of a strongly altered content in quinones [51] , [52] , [53] , [54] . Indeed , while wild-type animals contain endogenously synthesized UQ9 as well as a small amount of dietary bacterial UQ8 , clk-1 mutants contain only the dietary ubiquinone and no UQ9 . Here we found that clk-1 mutants have normal superoxide generation but enhanced overall ROS levels , which suggests that the antioxidant function of UQ9 is a crucial sink for mitochondrial ROS , whose absence appears to lead to an increase of overall ROS even in the absence of increase superoxide generation . eat-2 mutants are long-lived because of reduced food intake ( dietary restriction ) [55] . Although dietary restriction has been found to impinge on mitochondrial function in other systems , no changes in mitochondrial superoxide and overall ROS signals were observed . To determine how the elevated superoxide affects the lifespan of mutants , we treated worms with 10 mM of NAC and scored their survival ( Figure 2 and Table 1 ) . The treatment had no effect on the survival of the wild type ( Figure 2A ) , which shows that it is not toxic for lifespan at the concentration used . However , NAC treatment fully abolished the increased longevity of nuo-6 and severely limited that of isp-1 ( Figure 2B and 2C ) . The lesser effect on isp-1 is consistent with the larger increase of superoxide in these mutants ( Figure 1D ) , given that the effect of NAC is gradual ( 1 mM has less effect than 8 mM , which has less than 10 mM; Table S1 ) . At high concentration ( >10–15 mM ) NAC can be deleterious even on the wild type , but at the concentration used ( 10 mM ) NAC had no effect on the apparent health of the mutants , whose overall aspect after treatment was indistinguishable from that of the untreated worms ( Figure S2A ) . We have also quantified several phenotypes , including defecation , swimming , brood size , and post-embryonic development , after NAC treatment of the wild type and of nuo-6 , which is the mutant that is most sensitive to NAC ( 10 mM NAC completely abolishes its increased longevity ) . Treatment with 1 mM vitamin C also significantly shortened the lifespan of both isp-1 and nuo-6 mutants without affecting the wild type ( Table S1 ) . Most effects of NAC were quite small ( Figure S2B–E ) , except on the post-embryonic development of the wild type ( Figure S2C ) . Furthermore , for defecation , brood size , and post-embryonic development , the effect of NAC on the mutant produced a change in the same direction as on the wild type but of a lesser extent . Only for swimming is the effect greater on the mutant . But the effect consists of swimming faster after NAC treatment and thus bringing the mutant phenotype closer to the wild-type . We conclude that there is little evidence of an indirect deleterious effect of NAC . NAC had only a moderate effect on the lifespan of the insulin-signaling daf-2 mutants ( Figure 2E ) , suggesting that only a small part of the increased longevity of these mutants requires elevated mitochondrial superoxide . However , NAC fully abolished the increased lifespan of sod-2 mutants ( Figure 2F ) , suggesting that , although increased generation of superoxide and other ROS as detected by our techniques were not significantly altered in these mutants , their increased lifespan depends on an elevation of superoxide or some other ROS . NAC did not shorten the lifespan of clk-1 mutants at 10 mM ( Figure 2D ) , or even at 15 mM ( Table S1 ) , indicating that ROS metabolism is relatively irrelevant to the aging phenotype of these mutants . The effect of NAC on the lifespan of eat-2 could not be scored because NAC treatment rendered the animals unable to lay their eggs and they died from internal hatching at a young age . The origin of this effect is unknown . We also could not score the effect of NAC on RNAi-treated worms because 10 mM NAC was excessively damaging to the dsRNA-producing bacterial strain ( HT115 ) . To determine whether an elevation in mitochondrial superoxide generation is sufficient to increase lifespan , we used the superoxide generator PQ . Treatment of C . elegans with high concentration of PQ ( >0 . 2 mM ) is severely deleterious . We thus first tested the ability of PQ to increase ROS damage in the animals at a very low concentration ( 0 . 1 mM ) . We found that this treatment indeed measurably increased the level of oxidative damage to proteins at the young adult stage as assessed by determination of protein carbonylation ( Figure 3A ) and increased the expression of both the main cytoplasmic ( SOD-1 ) and the main mitochondrial ( SOD-2 ) superoxide dismutases ( Figure 3B and 3C ) . We then tested whether PQ could increase the lifespan of the wild type at three different concentrations ( 0 . 05 , 0 . 1 , and 0 . 2 mM ) and found that at all three concentrations both the mean and maximum lifespan were increased , with a maximal effect at 0 . 1 mM ( Figures 3D and 4A , and Tables 1 and S1 ) . The effect of 0 . 2 mM was less pronounced than that of 0 . 1 mM and similar to that of 0 . 05 mM , likely because at 0 . 2 mM a toxic effect starts to balance the pro-longevity effect . The effect does not depend on the exact chemical structure of paraquat , as benzyl-viologen , a compound with similar activity as PQ but structurally different , also increases lifespan ( Table S1 ) . A small effect of the prooxidant juglone under different conditions has also been documented previously [56] . The effect did not depend on an effect of PQ on the E . coli ( OP50 ) food source , as the effect was also observed with heat-killed cells ( Table S1 ) . Finally , the effect was not confined to development or adulthood as PQ prolongs lifespan whether provided only during adult lifespan or only during development ( Table S1 ) . PQ treatment failed to significantly prolong the lifespan on nuo-6 and isp-1 mutants ( Figure 4B and 4C , and Tables 1 and S1 ) . This experiment is equivalent to genetic epistasis analysis and suggests that nuo-6 , isp-1 , and PQ increase lifespan by the same mechanism . It also suggests that the maximum level of lifespan extension that can be obtained by increasing mitochondrial superoxide generation is already reached in these two mutants and further increase of superoxide generation through PQ treatment cannot increase lifespan further . This was not due to a resistance of these mutants to PQ as 0 . 2 mM PQ shortened the lifespan of the two mutants ( Table S1 ) . sod-2 mutants , whose longevity is suppressed by NAC , are not as long-lived when untreated as wild type animals that are treated with PQ . However , treatment with PQ makes the sod-2 mutants live as long as wild type animals treated with PQ ( Figure 4G ) . This absence of additivity suggests that the longevity of sod-2 mutants is indeed due to a small increase in superoxide , as expected from the function of SOD-2 , and the suppressing effect of NAC on the mutant lifespan . In contrast to what we observed with nuo-6 , isp-1 , and sod-2 , PQ treatment dramatically enhanced the lifespan of clk-1 and eat-2 mutants , significantly beyond the longevity increases induced by the mutations alone or by PQ applied to the wild type ( Figure 4D and 4E , and Tables 1 and S1 ) . This indicates that the effects of these mutations and the effect of superoxide are mechanistically distinct and additive , as expected from the finding that clk-1 and eat-2 mutants did not show increased mitochondrial superoxide levels ( Figure 1F ) and that the lifespan of clk-1 mutants could not be shortened by NAC treatment ( Figure 2D ) . PQ treatment had only a small lifespan-lengthening effect on daf-2 ( Figure 4F , and Tables 1 and S1 ) , which is consistent with the finding that daf-2 mutants already show a substantial increase in superoxide generation . We sought to determine whether the mutations and the PQ treatment had other common effects on mitochondrial function that could be responsible for the increased lifespans , besides elevation of superoxide levels . Work in other systems has suggested that increased mitochondrial biogenesis could impact lifespan positively [57] , [58] , [59] , and mitochondrial defects in C . elegans have been found to stimulate mitochondrial biogenesis , resulting in a denser mitochondrial network [13] . We have examined mitochondrial density in the two mitochondrial mutants and in PQ-treated worms with Mitotracker Red , which is specific to mitochondria in mammalian cells [60] , [61] , which stains worms uniformly , and whose staining fully overlaps with that of mitochondrially targeted green fluorescent protein ( GFP ) ( Figure S3 ) . We found that isp-1 and nuo-6 display a denser mitochondrial network , as expected ( Figure 5 ) . However , this was not observed in wild type worms treated with PQ ( Figure 5 ) , indicating that the mechanism by which the superoxide triggers longevity does not require increased mitochondrial biogenesis . We also tested the effects of PQ and NAC treatment on oxygen consumption and ATP levels in the wild type and in the two mitochondrial mutants ( Figure S4 ) . NAC treatment increased oxygen consumption in the wild type and in the mutants . This result uncouples oxygen consumption from lifespan as NAC has no effect on the lifespan of the wild type , and its effect on the oxygen consumption of isp-1 mutants is larger than on that of nuo-6 mutants , although its effect on aging is smaller . PQ had an effect only on nuo-6 , and it was small . Thus the effect of PQ on oxygen consumption also did not mirror its effect on lifespan . For ATP levels the only effect observed was a reduction by PQ of the elevated ATP levels that are observed in nuo-6 mutants . daf-2 mutants have elevated superoxide levels , and they are sensitive to NAC ( lifespan shortening by 15% ) . However , the level of superoxide in daf-2 appears not to be sufficient for a maximal effect as these mutants remain somewhat sensitive to PQ ( lifespan lengthening by 9% ) . To further study how superoxide plays a role in the lifespan of daf-2 we studied genes that function downstream of daf-2 . At least three genes are known to be required for the full lifespan extension of daf-2 , that is , daf-16 , aak-2 , and hsf-1 [62] , [63] , [64] . If one of these genes were necessary for an activity that mediates the small effect of PQ on daf-2 mutants , PQ should not be able to prolong the lifespan of mutants of such a gene . In fact , however , we found that PQ prolonged the lifespan of all three mutants tested ( Table 1 ) . The lifespan increase upon PQ treatment of daf-16 ( 35% increase ) and aak-2 ( 29% increase ) is not as large as upon treatment of the wild type ( 58% increase ) . This suggests that part but not all of the lifespan increase determined by superoxide requires daf-16 and aak-2 . These findings are consistent with the observations that the lifespan extension provided by nuo-6 and daf-2 ( e1370 ) are only partially additive ( Table S1 ) , similarly to what was found previously for isp-1 and daf-2 [12] , and that elimination of daf-16 partially shortens the lifespan of isp-1 [12] . We also tested the sensitivity to PQ of mutants that are diagnostic of a variety of pathways of aging . In particular mutants of genes that , based on their known functions in C . elegans or that of their homologues in other systems , might encode the targets of superoxide signaling or be otherwise necessary for implementing superoxide signaling . The c-Jun N-terminal kinase 1 ( JNK-1 ) is involved in stress responses in vertebrate cells and is a positive regulator of DAF-16 that acts in parallel to the effect of daf-2 on daf-16 [65] . We treated jnk-1 ( gk7 ) mutants with PQ and obtained a particularly large lifespan increase ( Table 1 ) . Although it is not clear what activities lie upstream of jnk-1 in C . elegans nor whether it has other targets than daf-16 , its activity does not appear necessary for the effect of superoxide . The transcription factor SKN-1 defends against oxidative stress by mobilizing the conserved phase II detoxification response and can delay aging independently of DAF-16 [66] . Although PQ induces oxidative stress and induces enzymes that protect from oxidative stress ( Figure 3 ) , it was still able to prolong the lifespan of skn-1 ( zn67 ) mutants ( Table 1 ) , indicating that skn-1 does not act downstream of superoxide . wwp-1 encodes a conserved E3 ubiquitin ligase that is necessary for lifespan extension by dietary restriction [67] . Treatment of wwp-1 ( ok1102 ) with PQ prolonged lifespan of these mutants , which is consistent with our finding that PQ can considerably extend the lifespan of eat-2 mutants ( Figure 4E ) . This confirmed that the lifespan increase produced by the superoxide increase in mitochondrial mutants is distinct from the mechanisms that support the lifespan effects of dietary restriction [14] . hif-1 encodes a worm homologue of the vertebrate hypoxia inducible factor 1α ( HIF-1α ) , a transcription factor involved in a number of protective mechanisms . In C . elegans hif-1 is necessary for a lifespan pathway that involves proteolysis and that is distinct from insulin signaling [68] and has also been involved in the dietary restriction pathway [69] . In vertebrates HIF-1α is positively regulated by mitochondrial ROS [34] , [35] , which would make it an interesting candidate to mediate the effects of superoxide . However , PQ was fully capable of increasing the lifespan of the hif-1 mutants ( Table 1 ) . Several of the genes whose mutants remain sensitive to PQ , including daf-16 , have been involved in stress responses , including oxidative stress , yet they do not seem necessary for the effect of PQ . Similarly we have shown previously that although the expression of SOD-1 and SOD-2 are elevated in isp-1 ( qm150 ) mutants , the elevation is not necessary for the extended lifespan of these mutants [24] . nuo-6 ( qm200 ) mutants also show elevated SOD-1 and SOD-2 expression [14] , but this too is unnecessary for the longevity of the mutants , as RNAi against sod-1 an sod-2 , which we have shown to be efficient in reducing enzyme levels [24] , does not shorten the lifespan of nuo-6 mutants ( Figure S5 ) . We conclude that the mitochondrial mutants protect from an aspect of the aging process that has not yet been studied through mutants that affect stress . In addition , our observations suggest that the lifespan effect we observed is not hormetic , as neither superoxide-detoxifying enzymes , nor the regulatory factors that are involved in protection from oxidative stress , are crucially implicated .
We have shown previously that mutations in isp-1 and nuo-6 prolong lifespan by a common mechanism [14] . Using direct measurement of ROS and superoxide we find here that this mechanism involves an increase in mitochondrial superoxide generation that is necessary and sufficient for the longevity of these mutants . As ROS , including superoxide [70] , [71] , [72] , are known to be intracellular messengers , the increased superoxide might trigger a signal transduction pathway that ultimately results in changes in nuclear gene expression [23] . Superoxide is highly reactive and could trigger such a signal by modifying proteins in the mitochondria or in the nearby cytosol after having escaped from the mitochondria through an appropriate channel [73] , [74] . Although no superoxide sensor has yet been identified , a similar type of mechanism , in which a highly reactive , quickly diffusing , molecule modifies a signal transduction protein , has been evidenced for nitric oxide ( NO ) , which covalently and permanently modifies guanylyl cyclases . Similarly , hydrogen peroxide ( H2O2 ) , the product of superoxide dismutation , can inactivate phosphatases involved in signal transduction . Future work will aim at using forward and reverse genetic screens in C . elegans to uncover the molecular machinery that reacts to the superoxide signal , as well as the transcription factors that are needed to regulate nuclear gene expression in response to the pathway's activation . In addition , the pattern of gene expression that results in increased lifespan in these mutants could be defined very specifically by identifying changes in the gene expression patterns that are common to isp-1 , nuo-6 , and PQ treatment and that are suppressed by treatment with NAC . A number of studies in C . elegans have explored hormesis by treating animals with sub-lethal but clearly deleterious treatments for a short period of time and observing subsequent prolongation of lifespan [75] . These hormetic effects are different from what we have observed and describe here , as both the genetic mutations and the very low level PQ are present throughout life and as only a part of the effect we observe might require the insulin signaling pathway . Furthermore , although in nuo-6 and isp-1 mutants the expression levels of the superoxide dismutases SOD-1 and SOD-2 are elevated , likely in response to the elevated superoxide generation , and as one expects in the hormetic response , these elevations are not necessary for the lifespan prolongation of nuo-6 ( Figure S5 ) or isp-1 [24] . CLK-1 is a mitochondrial protein that is required for ubiquinone biosynthesis and its absence affects mitochondrial function [50] , although it could potentially affect many other processes as ubiquinone is found in all membranes . Furthermore , ubiquinone is both a prooxidant as co-factor in the respiratory chain and an anti-oxidant . Interestingly , the mechanism of lifespan prolongation induced by clk-1 appears to be entirely distinct from , but particularly synergistic with , that induced by elevated superoxide . Indeed , clk-1 mutants do not show elevated superoxide generation and are not affected by NAC . Furthermore , although double mutant combinations of clk-1 with nuo-6 and isp-1 are not viable ( unpublished data ) the lifespans of clk-1 mutants treated with PQ ( Figure 4D ) , or of sod-2;clk-1 mutants [26] , or of clk-1;daf-2 mutants [76] are much greater than expected from simple additivity of the effects of individual mutations or treatments . Studies in yeast [77] and in worms [78] have suggested that an increase in ROS from mitochondria might also be important in triggering the lifespan extension produced by glucose restriction . However , our results here with an eat-2 mutation , one of the ways in which global dietary restriction can be produced in worms , as well as with a wwp-1 and hif-1 , which may function downstream of dietary restriction , did not reveal an involvement of superoxide signaling , providing further evidence for a distinction between the mechanisms of glucose restriction and dietary restriction . It remains possible , however , that DR could lead to superoxide or ROS production when it is induced by other methods than the use of an eat-2 mutant , as it is well documented that different types of DR induce different molecular mechanisms [79] . One question that our current experiments do not address is whether the mitochondrial dysfunction in the mutants , or the effect of PQ , is necessary in every tissue in order to increase longevity . There are indications for both the insulin signaling pathway mutants [80] , [81] and dietary restriction [67] , [82] that the entire effect might be mediated by action in particular cells that influence the physiology of the whole organism . Similarly , the presence or absence of the germline is sufficient for a dramatic effect on lifespan [83] . For mitochondrial dysfunction the question could be addressed in the future by mosaic analysis and by purifying and analyzing mitochondria from specific tissues using our flow cytometry technique to purify mitochondria expressing GFP in a tissue-specific manner . The oxidative stress theory of aging has been one of the most acknowledged theories of aging for the simple reason of the strikingly good correlation between the levels of oxidative stress and the aged phenotype [8] . A number of recent results in worms and in mice , however , have suggested that oxidative stress cannot be the cause of aging [24] , [25] , [26] , [30] . Our findings suggest a conceptual framework for why oxidative stress and the aged phenotype are so tightly correlated [31] . In this model mitochondria , like the rest of the cell , sustain a variety of age-dependent insults ( not only and not even principally from oxidative stress ) that trigger an increase in superoxide , which acts as a signal that induces general protective and repair mechanisms . However , aging in most animals is clearly irreversible , indicating that the protective mechanisms , which must have evolved to control damage in young organisms , are unable to fully prevent the accumulation of age-related damage . Thus , as superoxide is a reactive molecule as well as a signal , and as age-dependent damage cannot be fully reversed , it is possible that at high ages the chronically elevated superoxide will participate in creating some of the damage itself . This could explain the strong tendency for aged animals to have high oxidative stress and high oxidative damage , although it does not imply that ROS cause aging or even that they are a major source of age-dependent damage . In this model , the nuo-6 and isp-1 mutations lead to increased longevity because they turn on the stress signal prematurely and thus slow down the entire process .
Eggs were placed on plates at 20°C and left for 1 h to hatch . Larvae that had hatched during that period were placed onto fresh plates and monitored once daily until death . The animals were transferred once daily while producing eggs to keep them separate from their progeny . Animals were scored as dead when they no longer responded with movement to light prodding on the head and tail . Missing worms and worms that have died because of internal hatching ( bagging ) were replaced from a backup group . Survival was scored every day . Drugs were added into NGM media from a high concentration stock solution ( 500 mM for NAC , 1 M for PQ , and 500 mM for vitamin C ) before pouring of the plates . Plates were made fresh each week . Gravid adult worms were transferred from normal NGM plates to drug plates and left to lay eggs for 3 h . With each transfer of worms a substantial amount of bacteria was also transferred onto the new plates . The progeny was then scored for different phenotypes . All dyes except MitoSox were diluted in DMSO at high concentration ( all at 5 mM except H2DCFDA , which is at 10 mM ) and frozen at −20°C as a stock . MitoSox was prepared fresh at 5 mM for each use . Before staining stocks were diluted in M9 buffer at a 1∶1000 dilution . Young adult worms were transferred into staining solution and stained for 20 min . Worms were mounted on a thick layer of half-dried agar pad on microscopic glass slides and then subjected to confocal microscopy ( Zeiss LSM 510 Meta ) . Pictures were taken by Zeiss LSM Imaging software and analyzed by Volocity V4 . 0 software . Five young adult worms ( 1st day of adulthood ) were placed into 0 . 25 µl M9 buffer in a 0 . 5 µl sealed chamber at 22°C . A fiber optical oxygen sensor ( AL300 FOXY probe from Ocean Optics ) was inserted into this chamber and oxygen partial pressure was monitored for 15 to 30 min . Oxygen consumption measured in this way was normalized to body volume . For this worms were photographed at each measurement day under a binocular microscope and their cross-section was calculated with ImageJ software . Worm volume was determined by the formula: volume ( nl ) = 1 . 849 • 10–7 ( nl/µm3 ) • area 1 . 5 ( µm3 ) [84] . After RNAi treatment , 100 young adult worms of each genotype were picked , lysed in 2× loading buffer , and subjected to electrophoresis in 12% SDS–polyacrylamide gels ( SDS–PAGE ) , and then blotted onto nitrocellulose membrane ( Bio-Rad ) . After applying primary antibody ( 1∶1000 , rabbit polyclonal antibody against worm SOD-1 or SOD-2 ) and secondary antibody ( 1∶10 , 000 mouse anti-rabbit IgG , Invitrogen ) , the membranes were incubated with the ECL plus detection reagent ( Amersham Biosciences ) and scanned using a Typhoon trio plus scanner . Band densities were analyzed by ImageQuant TL V2003 . 03 . For fluorescence activated cell sorting [37] , adult worms grown on large NGM plates were collected and washed 3 times with M9 buffer . Worms were then suspended in 5× isolation buffer ( 200 mM mannitol; 120 mM sucrose; 10 mM Tris; 1 mM EGTA; pH 7 . 4 ) and set on ice . Worms were broken up with a 5 ml glass-glass homogenizer and centrifuged at 600 g for 10 min , the supernatant was collected and re-centrifuged at 7 , 800 g for 10 min , and the pellet was washed once with isolation buffer and then suspended in isolation buffer and kept on ice . Different dyes were added from stocks into the analysis buffer ( 250 mM sucrose; 20 mM MOPS; 100 uM KPi; 0 . 5 mM MgCl2; 1 uM CsA pH 7 . 0 ) at a 1∶1000 dilution before staining . 100 µl of mitochondria was added to 900 µl of analysis buffer with dye and substrate and incubated for 1 h at room temperature . Mitochondria were recollected by 7 , 800 g centrifugation and then suspended in 500 µl analysis buffer . A FACSCalibur instrument equipped with a 488 nm Argon laser and a 635 nm red diode laser ( Becton Dickinson ) was used . Data from the experiments were analyzed using the CellQuest software ( Becton Dickinson ) . To exclude debris , samples were gated based on light-scattering properties in the SSC ( side scatter ) and FSC ( forward scatter ) modes , and 20 , 000 events per sample were collected , using the “low” setting for sample flow rate ( Figure S1 ) . 200 age-synchronized young adult worms were collected in M9 buffer and washed three times . Worm pellets were treated with three freeze/thaw cycles and boiled for 15 min to release ATP and destroy ATPase activity , and then spun at 4°C and 11 , 000 g for 10 min . ATP contents were measured with a kit ( Invitrogen , Carlsbad , California , USA; Cat: A22066 ) . The ATP content value was then normalized to the soluble protein level of the same preparation , measured with the protein assay from Bio-Rad . Mitotracker green ( Invitrogen M7514 ) stock concentration 5 mM; H2DCFDA ( Invitrogen D399 ) stock concentration 10 mM; Mitotracker red ( Invitrogen M7512 ) stock concentration 5 mM . | An unequivocal demonstration that mitochondria are important for lifespan comes from studies with the nematode Caenorhabditis elegans . Mutations in mitochondrial proteins such as ISP-1 and NUO-6 , which function directly in mitochondrial electron transport , lead to a dramatic increase in the lifespan of this organism . One theory proposes that toxicity of mitochondrial reactive oxygen species ( ROS ) is the cause of aging and predicts that the generation of the ROS superoxide should be low in these mutants . Here we have measured superoxide generation in these mutants and found that it is in fact elevated , rather than reduced . Furthermore , we found that this elevation is necessary and sufficient for longevity , as it is abolished by antioxidants and induced by mild treatment with oxidants . This suggests that superoxide can act as a signal triggering cellular changes that attenuate the effects of aging . This idea suggests a new model for the well-documented correlation between ROS and the aged phenotype . We propose that a gradual increase of molecular damage during aging triggers a concurrent , gradually intensifying , protective superoxide response . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"developmental",
"biology/aging"
] | 2010 | A Mitochondrial Superoxide Signal Triggers Increased Longevity in Caenorhabditis elegans |
The Ras-GAP SH3 domain–binding proteins ( G3BP ) are essential regulators of the formation of stress granules ( SG ) , cytosolic aggregates of proteins and RNA that are induced upon cellular stress , such as virus infection . Many viruses , including Semliki Forest virus ( SFV ) , block SG induction by targeting G3BP . In this work , we demonstrate that the G3BP-binding motif of SFV nsP3 consists of two FGDF motifs , in which both phenylalanine and the glycine residue are essential for binding . In addition , we show that binding of the cellular G3BP-binding partner USP10 is also mediated by an FGDF motif . Overexpression of wt USP10 , but not a mutant lacking the FGDF-motif , blocks SG assembly . Further , we identified FGDF-mediated G3BP binding site in herpes simplex virus ( HSV ) protein ICP8 , and show that ICP8 binding to G3BP also inhibits SG formation , which is a novel function of HSV ICP8 . We present a model of the three-dimensional structure of G3BP bound to an FGDF-containing peptide , likely representing a binding mode shared by many proteins to target G3BP .
The Ras-GAP SH3 domain–binding proteins ( G3BP ) are multifunctional RNA-binding proteins , present in two forms , G3BP-1 and G3BP-2 ( here collectively referred to as G3BP ) . They have a well-described importance in mediating the formation of RNA stress granules ( SG ) , both in cells exposed to environmental stress and viral infections [1 , 2] . SGs are formed when translation initiation is compromised after phosphorylation of eukaryotic initiation factor eIF2α [3] or inhibition of eIF4A [4] . The assembly of SGs allows for rapid redirection of translation to stress response mRNAs or , in the case of viral infection , for inhibition of viral gene expression . The G3BP proteins possess RNA recognition motifs ( RRM ) , which , together with protein/protein interaction domains , are required for SG induction [2] . The N-terminus of G3BP comprises a nuclear transport factor 2 ( NTF2 ) -like domain [5] , which is likely involved in dimerization [5 , 6] , but little is known about the functional consequences of such dimerization . The G3BP NTF2-like domain forms complexes with a number of cellular proteins such as ubiquitin-specific protease 10 ( USP10 ) , caprin-1 and OGFOD-1 [7–9] . G3BP-binding regulates the activity of USP10 , a predominantly cytoplasmic deubiquitinating enzyme ( DUB ) [8] which stabilizes several important proteins including the cystic fibrosis transmembrane conductance regulator ( CFTR ) [10] , the tumor suppressor p53 [11] , the autophagy regulator Beclin-1 [12] , the sirtuin family histone deacetylase SIRT6 [13] , the NF‐kB essential modulator ( NEMO/IKKγ ) [14] and the transporter associated with antigen processing ( TAP1 ) [15] . The G3BP binding region of USP10 is found within its N-terminal 76 residues [16] , and this interaction inhibits the DUB activity [8 , 17] . SGs are induced by many virus infections and in turn , viruses have evolved many countermeasures , often targeting G3BP [18] . SG assembly in poliovirus infection is inhibited by cleavage of G3BP between residues Q325 and G326 by the viral 3C protease [19] separating the NTF2-like and RRM domains and leading to the formation of compositionally distinct SGs , lacking G3BP [20] . For some viruses , G3BP is recruited to foci of viral protein accumulation and may be important for efficient completion of the viral life cycle . In vaccinia virus ( VV ) -infected cells , G3BP is recruited to the cytoplasmic viral factories [21] . However , it has also been reported to have an antiviral role in VV infection [22] . Likewise , G3BP has been implicated as a potential component of the hepatitis C virus ( HCV ) replication complex [23] and may play an important role in virus assembly [24] . We and others have shown that the G3BP NTF2-like domain is directly bound by L/ITFGDFD repeat motifs in the C-termini of non-structural protein ( nsP ) 3 of the Old World alphaviruses , including Semliki Forest virus ( SFV ) and chikungunya virus ( CHIKV ) [25–28] . Subsequent sequestration of G3BP to foci of viral protein accumulation renders the infected cells unable to assemble SGs , despite sustained high levels of eIF2α phosphorylation [28 , 29] . In this work , we set out to precisely define the characteristics of the G3BP-binding motif in the Old World alphavirus nsP3 . We found that the core binding motif consists of two phenylalanine residues separated by a glycine and an aspartate residue ( FGDF ) . Mutation of either of these residues prevented binding to either G3BP-1 or G3BP-2 . We also find FGDF motifs at the N-terminus of USP10 and at the C-terminus of herpes simplex virus ( HSV ) -1 protein ICP8 , and confirmed that these motifs similarly bind G3BP and block SG induction . Finally , we generated a molecular model of the G3BP/FGDF peptide interaction and validated the model by site-directed mutagenesis . These results reveal the mechanism by which at least two pathogenic viruses target G3BP . Moreover , they also provide a molecular basis for the regulation of USP10 activity by assembly of the inhibitory USP10/G3BP complex .
The C-terminal L/ITFGDFD repeat motifs , which constitute the G3BP-binding site of nsP3 , are well conserved in the Old World alphaviruses [30] with particularly strong conservation of the phenylalanine residues at positions 3 and 6 of the seven residue motif and of the glycine residue at position 4 . In order to identify essential residues in the SFV sequence for G3BP binding , we constructed mutants in which each residue from T2 to D7 in both repeats was exchanged for alanine ( Fig . 1A ) . These were constructed in the context of the pEGFP-nsP3-31 construct , containing amino acids 447–477 of SFV nsP3 fused to the C-terminus of EGFP . The empty pEGFP-C1 vector , encoding EGFP with a C-terminal extension of 21 residues was used as a control . We previously demonstrated that EGFP-nsP3-31 and an analogous EGFP fusion protein containing amino acids 475–523 of the CHIKV nsP3 efficiently bind G3BP in transient transfection experiments [27 , 28] . BHK cells were transfected with vectors encoding EGFP alone , EGFP-nsP3-31 or each of the alanine mutants . Lysates were immunoprecipitated with G3BP-1 antibody and immunoblotted with antibodies against G3BP-1 , GFP or actin ( Fig . 1B , upper panels ) . As expected , EGFP-nsP3-31 , containing the wild type ( wt ) nsP3 sequence , efficiently bound G3BP-1 , but EGFP alone did not . The construct carrying a mutation of the threonine residue T2 to an alanine exhibited weak but readily detectable binding to G3BP . However , mutation of either of the residues F3 , G4 or F6 completely disrupted binding . Finally , while mutation of the aspartate residues at D5 to alanine substantially reduced binding , mutation of D7 had no effect . Lysates were also immunoprecipitated with anti-GFP and immunoblotted for G3BP-1 , GFP or actin ( Fig . 1B , middle panels ) , with similar results . Furthermore , GFP-immunoprecipitates were immunoblotted for G2BP-2 ( S1 Fig . ) . These results show that both alternatively-spliced isoforms of G3BP-2 ( a and b ) interact with EGFP-31 constructs in a similar manner to G3BP-1 . We conclude from these results that residues F3 , G4 and F6 are essential for G3BP-binding and that T2 and D5 contribute significantly . Previously we have shown that in Old World alphavirus-infected cells , the C-terminal repeat sequences of nsP3 bind and sequester both G3BP-1 and G3BP-2 into foci of viral protein accumulation including cytopathic vacuoles ( CPV ) [27 , 28 , 31] . In order to test the effects of the non-G3BP-binding mutations on the life cycle of SFV , we mutated both nsP3 residues F451 and F468 to alanines in the infectious clone of SFV ( corresponding to the mutation F3A in the two ( L/I ) TFGDFD motifs , hereafter referred to as SFV-F3A ) . Previously , we showed that SFV-Δ789 , lacking nsP3 residues 449–472 , including the G3BP-binding sites displayed slower processing of the P34 precursor [28] . To determine if SFV-F3A also displays slow processing phenotype , we infected BHK cells with wt SFV , SFV-F3A or SFV-Δ789 and compared levels of P34 . Lysates from WT SFV or SFV-F3A-infected cells did not contain detectable levels of P34 ( S2A Fig . ) , and we concluded that the F3A mutations do not result in defective P34 processing . To demonstrate that nsP3-F3A does not bind G3BP-1 in the context of a viral infection , we infected BHK cells with wt SFV or SFV-F3A at MOI 10 . Lysates were immunoprecipitated with G3BP-1 antibody and analyzed by immunoblotting for nsP3 , G3BP-1 and actin . As expected , nsP3 from wt SFV-infected cells coprecipitated with G3BP-1 , but nsP3 F3A did not ( Fig . 2A ) . To determine whether SFV-F3A nsP3 colocalized with G3BP , cells were infected at low MOI with wt SFV or SFV-F3A . Cells were fixed and stained for SFV nsP3 , G3BP-1 and TIA-1 , another SG marker . As expected , wt SFV nsP3 colocalized well with G3BP-1 , with very little colocalization of G3BP-1 with TIA-1 ( Fig . 2B ) . However , nsP3 in SFV-F3A infected cells did not colocalize with G3BP-1 , confirming immunoprecipitation data indicating that nsP3 F3A does not interact with G3BP-1 . Instead , SFV-F3A infected cells often contained foci of G3BP-1 and TIA-1 co-staining , suggesting longer persistence of SGs in cells infected with the SFV-F3A mutant . Previously we demonstrated that at late times in SFV infection , when most of the infection-induced SG have been disassembled , SGs cannot be re-induced with phospho-eIF2α- or eIF4A-dependent stress inducers ( sodium arsenite or pateamine A ( Pat A ) , respectively ) because G3BP is sequestered by nsP3 [28 , 29] . We next tested whether cells infected with wt SFV or SFV-F3A could mount an SG response to an exogenous stress inducer . Cells were infected with wt SFV or SFV-F3A and stressed with Pat A at 7 hpi for 1 h before fixation and staining for TIA-1 to detect SGs and nsP3 to detect SFV infection ( Fig . 2C ) . Pat A was used since sodium arsenite has little effect in SFV-infected cells as these already display sustained eIF2α phosphorylation [28] . Approximately 100% of mock-infected cells responded to Pat A by forming SGs . After infection with wt SFV , 5 . 2% of cells had SGs , and this number was not significantly altered by Pat A treatment , confirming that wt SFV–infected cells cannot mount a stress response to secondary stress signals . SFV-F3A–infected cells had a significantly higher number of SG-positive cells than wt SFV–infected cultures ( 18 . 4% ) , and this proportion was increased to 49 . 2% by the Pat A treatment , indicating that in the absence of nsP3/G3BP interaction , infection-induced SGs persist longer and additional SGs can be stress-induced in SFV-F3A-infected cells . To determine whether the specific ablation of G3BP recruitment affects viral replication , we performed single-step and multi-step growth curves comparing wt SFV and SFV-F3A in MEFs . These data revealed that SFV-F3A propagated to titers between 1 . 5 and 2 orders of magnitude lower that wt SFV after both low and high MOI infection ( Fig . 2D ) indicating that blocking the nsP3/G3BP interaction by point mutation of the N-terminal phenylalanine residue in both FGDF motifs attenuates viral infection . Single and multi-step growth curve experiments were also performed in BHK cells with similar results ( S2B Fig . ) . Finally , to address whether the F3A mutations in SFV-nsP3 affect other functions than SG inhibition by G3BP sequestration , we performed viral growth curves in eIF2α-AA MEF cells [32] . These cells contain a mutation at the phosphorylation site ( S51A ) in eIF2α gene , such that eIF2α cannot be phosphorylated and SGs are not induced . SFV-F3A replication was very similar to WT SFV in those cells ( Fig . 2E ) . This result strongly suggests that the attenuation of SFV-F3A in WT cells is due to the action of a G3BP-dependent process downstream of eIF2α phosphorylation , most likely SG formation . Hence , inhibition of SG formation ( in response to infection-induced eIF2α phosphorylation ) is probably the primary function of the FGDF motifs in Old World alphavirus nsP3 . We have previously demonstrated that at late times in infection with SFV , a large proportion of cellular G3BP is recruited to sites of viral protein accumulation [28] . We wondered whether this interaction displaces USP10 , a cellular binding partner of G3BP that also binds to the NTF2-like domain [8 , 25] . When we determined the localization of USP10 in SFV-infected MEFs , we did not detect any significant overlap with G3BP ( S3 Fig . ) . This suggests that USP10 is excluded from the complex with G3BP by the nsP3 interaction , possibly by competition for the same site on G3BP . Recently , the G3BP-binding domain of USP10 was shown to be located within the N-terminal 76 amino acid residues [16] . Having shown that the G3BP-binding motif of nsP3 consists of two FGDF motifs , we asked whether a similar motif exists in the G3BP-binding region of USP10 . When we aligned the sequences of the N-terminal 80 residues of USP10 from several species ( Fig . 3A ) , we identified several well conserved phenylalanine residues ( F10 , F13 , F18 , F21 , F22 and F59 in the human sequence ) , as well as a conserved FG ( D/E ) F motif at residues 10–13 . We observed particularly high species conservation in the N-terminal 40 residues . To determine whether this USP10 region binds G3BP , we fused the N-terminal 40 residues of human USP10 to EGFP ( EGFP-USP101–40 ) . BHK cells were transfected with vectors encoding EGFP alone or EGFP-USP101–40 . Lysates were immunoprecipitated with anti-GFP and immunoblotted with sera to G3BP-1 , G3BP-2 , GFP and actin ( Fig . 3B , left panels ) . Indeed , both G3BP proteins efficiently coprecipitated with EGFP-USP101–40 but not with EGFP alone . Lysates were also immunoprecipitated with G3BP-1 antibody to verify reciprocal coimmunoprecipitation of EGFP-USP101–40 ( Fig . 3B , middle panels ) . To determine whether this binding is dependent on the FGDF motif at residues 10–13 , a panel of mutated EGFP-USP101–40 variants was created in which each residue from F10 to F13 was substituted to alanine . Cells were transfected with vectors encoding wt EGFP , EGFP-USP101–40 or each variant . Lysates were immunoprecipitated with anti-GFP and immunoblotted with sera to G3BP-1 , G3BP-2 , GFP and actin ( Fig . 3C ) . Alanine substitution of F10 , G11 or F13 completely disrupted binding , while the D12A mutation allowed weak binding to both G3BP-1 and G3BP-2 . Similar results were obtained after G3BP-1 immunoprecipitation ( S4 Fig . ) . These results are remarkably similar to those obtained following mutagenesis of the G3BP-binding domain of SFV nsP3 ( Fig . 1 ) , suggesting that both proteins indeed bind G3BP via their FGDF-motifs . To test if the overexpression and binding of full-length USP10 to G3BP affects the SG response , BHK cells were transfected with EGFP alone , EGFP-USP10 wt or EGFP-USP10-F10A . Transfected cells were mock stressed or stressed with sodium arsenite for 1 h before fixation and staining for G3BP-1 and TIA-1 to detect SGs . Approximately 97% of the mock transfected cells and 71% of the EGFP transfected cells had SGs upon sodium arsenite treatment ( Fig . 3D ) . Overexpression of wt EGFP-USP10 efficiently blocked the ability of the cells to form SGs ( just 1% of cells showed SGs ) , whereas 59% cells transfected with the G3BP-nonbinding mutant EGFP-USP10-F10A responded to sodium arsenite stress with SGs . Representative images are presented in S5 Fig . . Taken together , the results in Fig . 3 show that USP10 acts as a negative regulator of G3BP-dependent SG formation and that this regulation is dependent on G3BP binding by the FGDF motif at residues 10–13 . In contrast to USP10 , nsP3 contains two FGDF motifs ( residues 451–454 and 468–471 ) . We therefore hypothesized that a minimal G3BP-binding motif could consist of only one FGDF motif flanked by surrounding residues , which would provide enough space between the two motifs to allow nsP3 to bind two G3BP molecules . Based on our finding that both G3BP-1 and G3BP-2 bind FGDF motifs , we further hypothesized that if EGFP-nsP3-31 binds two molecules of G3BP , it would bind any combination of G3BP-1 and -2 , with frequencies depending on the relative abundance of the two proteins and potential differences in affinity . Therefore , G3BP-2 should be detectable in nsP3 complexes immunoprecipitated with anti-G3BP-1 and vice versa , but USP10 should form a complex with only one molecule of either G3BP-1 or -2 ( Fig . 4A ) . To test this , BHK cells were transfected with vectors encoding EGFP , EGFP-USP101–40 or EGFP-nsP3-31 . When lysates were immunoprecipitated with anti-GFP , both G3BP-1 and G3BP-2 were detected in complex with EGFP-nsP3-31 and EGFP-USP101–40 ( Fig . 4B , left panels ) . However , when G3BP-1 was immunoprecipitated , G3BP-2 could be detected in EGFP-nsP3-31 , but not in EGFP-USP101–40 bound complexes ( Fig . 4B , middle panels ) . This suggests that the nsP3 sequence , containing two FGDF motifs , can form a ternary complex with both G3BP-1 and G3BP-2 simultaneously , while the USP10 sequence , containing only one FGDF motif , can bind either G3BP-1 or G3BP-2 , but not both . Next , we endeavored to confirm the 2:1 stoichiometry of nsP3:G3BP using biophysical in vitro measurements . We employed isothermal titration calorimetry ( ITC ) to determine the affinity , stoichiometry and thermodynamic signature of the interaction between the purified NTF2-like domain of human G3BP-1 , expressed in E . coli ( G3BP-NTF2 ) , and a peptide spanning residues 449–473 of SFV nsP3 ( nsP3-25 ) , containing two FGDF motifs . Injection of nsP3-25 into a G3BP-NTF2 protein solution resulted in concentration-dependent exothermic heat changes ( Fig . 4C , upper panel ) . The binding isotherm of the integrated heat changes was fitted using a simple one-to-one binding model , yielding an affinity value of 7 μM as well as 2 . 4 G3BP binding sites per nsP3-25 peptide molecule ( Fig . 4C , lower panel ) . This indicates that one peptide efficiently binds two G3BP molecules . The interaction affinity was 16-fold stronger than the previously determined affinity value of 115 μM for the interaction between a similar G3BP-NTF2 fragment and a DSGFSFGSK peptide [33] . Injection of a control peptide in which both FGDF motifs were exchanged for AGDA ( nsP3-25-mut ) only produced injection-related heat fluctuations at baseline level ( Fig . 4D ) . Engagement of two G3BP molecules with each nsP3-25 peptide was further supported by analytical size exclusion chromatography ( SEC ) analysis . It is known from crystal structures and SEC analysis that the NTF2-like domains of rat and human G3BP form homo-dimers [33] . G3BP-NTF2 alone was eluted with an apparent molecular weight of 30 kDa , slightly lower than the calculated molecular weight of the dimer ( 36 kDa ) , while pre-incubation of G3BP with a ten-fold molar excess of nsP3-25 peptide shifted the elution peak , corresponding to the formation of a 60 kDa complex ( Fig . 4E ) . Taking into account the 2:1 ratio for the G3BP-NTF2:nsP3-25 interaction as derived from ITC ( Fig . 4C ) , we suggest that the formed 60 kDa complex comprised four G3BP-NTF2 molecules ( two dimers ) and two nsP3-25 peptides . There was no evidence for higher-order oligomers . The crystal structures of the NTF2-like domain of G3BP-1 in its apo form and complexed with the nucleoporin ( Nup ) -derived peptide DSGFSFGSK have recently been determined [33] , revealing that the ligand is bound in an extended conformation within a long and deep groove on the surface of G3BP ( S6A Fig . ) . The peptide-binding site is amphipathic; while the base of the groove is highly hydrophobic , both walls lining the cleft of the protein and the positively charged N-terminus are polar with several basic residues ( R32 , K5 , K123 ) . Both phenylalanine residues of the peptide protrude towards well-defined pockets within the hydrophobic groove . While the side chain of residue F4 of the Nup-derived peptide is buried in a deep pocket formed by the G3BP-1 residues F15 , F33 and F124 , the side chain of residue F6 is localized in a shallower pocket formed by the G3BP-1 residues F124 , V11 and L10 [33] . Our binding and mutagenesis studies reveal that G3BP-1 binds approximately 16 times stronger to FGDF-containing peptides compared to the previously described interaction with the DSGFSFGSK peptide . Both these sequences contain two crucial phenylalanines , which we hypothesized to be bound similarly in both peptides . We assessed this by creating a molecular model in which the octapeptide LTFGDFDE was manually docked into the peptide-binding groove of G3BP-1 , using the crystal structure of the G3BP-NTF2 complexed with the DSGFSFGSK peptide as a template ( Figs . 5 and S6B ) . The conformational flexibility of the additional glycine residue within the FGDF motif would allow the phenylalanine side-chains to take similar orientations as in the SGFSF-peptide . Furthermore , the molecular model indicates that the negatively charged residues D5 , D7 and E8 in LTFGDFDE could interact with the positively charged side chains of residues K123 and K5 as well as with the positively charged N-terminal region of G3BP ( Fig . 5B ) . It should be noted that these two lysine residues are conserved in human , mouse and Xenopus G3BP-1 and are also present in human G3BP-2 ( S7 Fig . ) . The binding mode suggested by the molecular model of the LTFGDFDE/G3BP-1 complex is in agreement with our biochemical analyses , which showed that mutation of the conserved phenylalanine residues at positions 451 , 454 , 468 and 471 of nsP3 and 10 and 13 of USP10 eliminated G3BP-binding ( Figs . 1 and 3 ) . The model might also explain the elimination of binding of both nsP3 and USP10 to G3BP upon mutation of the glycine residue to alanine , which would restrict flexibility and thus hinder adequate positioning of the phenylalanine side-chains in the FGDF motifs . In order to validate the proposed binding model , residue F33 , localized within the hydrophobic pocket of G3BP and predicted to be proximal to residue F3 of the bound LTFGDFDE peptide , was mutated to tryptophan ( G3BP-F33W ) . Residue F33 is buried at the bottom of the peptide-binding cleft , and we hypothesized that its substitution to tryptophan would reduce the size of the pocket , thus hindering adequate positioning of the benzene ring of peptide residue F3 into the cleft ( Fig . 6A , upper panel ) . As a control , residue F124 was also mutated to tryptophan , since this phenylalanine residue is solvent-accessible and not localized within the binding groove ( Fig . 6A , lower panel ) . Notably , F33 is conserved in human , mouse , Xenopus and Aedes mosquito G3BP-1 , while F124 conserved in human , mouse and Xenopus , but a tyrosine residue in the Aedes sequence ( S7A Fig . ) . Both are conserved in human G3BP-2 ( S7B Fig . ) . To determine whether mutation of these residues affects binding of G3BP with FGDF motifs , nsP3 ( tagged with a biotin acceptor peptide , BAP ) was co-expressed with either pEGFP-C1 , pEGFP-G3BP-wt , pEGFP-G3BP-F33W or pEGFP-G3BP-F124W and binding was analyzed by immunoprecipitation with anti-GFP ( Fig . 6B ) . Consistent with our structural model , EGFP-G3BP-wt and the F124W variant efficiently bound to nsP3-BAP , while EGFP-G3BP-F33W did not detectably interact . Similarly to results with cotransfected nsP3 , endogenous USP10 was also coimmunoprecipitated with EGFP-G3BP-wt and EGFP-G3BP-F124W but not with EGFP-G3BP-F33W ( Fig . 6C ) . The extreme N-terminus of G3BP ( residues 1–11 ) forms a large part of the hydrophobic pocket for positioning of the peptide residue F6 and also contains a lysine residue ( K5 ) , possibly interacting with the negative charges of acidic residues downstream of the FGDF motif . The model predicts that a truncated version of G3BP ( Δ1–11 ) would not be capable of binding the peptide ( Fig . 6D ) . To evaluate this , 293T cells were transfected with pEGFP-C1 , pEGFP-G3BP-wt or pEGFP-G3BP-∆1–11 and analyzed by immunoprecipitation with anti-GFP and immunoblotting for USP10 , GFP or actin ( Fig . 6E ) . Endogenous USP10 coprecipitated with EGFP-G3BP-wt but not with EGFP-G3BP-∆1–11 . Taken together , the results in Fig . 6 strongly support our structural model of the G3BP/FGDF complex . Next , we asked whether the FGDF motif , mediating G3BP binding , is present in proteins other than SFV nsP3 and USP10 . To identify such proteins , we searched the UniProtKB human and virus databases for all proteins containing FGxF motifs , where x = D , E or S , and also containing at least two acidic residues within the downstream 5 residues ( as in both SFV nsP3 and USP10 ) . Glutamic acid was permitted in the third position due to its chemical similarity to aspartic acid , while serine was permitted since Sindbis virus nsP3 , shown by several researchers to bind G3BP-1 [25 , 26] , contains a serine at that position . We identified 34 human ( S1 Table ) and 32 viral ( S2 Table ) sequences that meet these criteria . In one of those proteins , the infected cell protein ( ICP ) 8 of herpes simplex virus ( HSV ) -1 , the FGDF motif is located between residues 1144 and 1147 in the C-terminal region of the 1196 aa-long protein that is not predicted to adopt any specific secondary structure [34] and contains 3 acidic residues within the downstream 5 residues . This is reminiscent of the context of the FGDF motifs in Old World alphavirus nsP3 , an otherwise quite different protein . ICP8 is a single stranded DNA binding protein that is expressed in the lytic cycle of herpes simplex viral replication . It is one of seven viral proteins that are necessary for viral DNA replication [35] . The FGDF motif in ICP8 is conserved in several HSV-1 strains and also in HSV-2 ( S2 Table ) . To determine whether ICP8 binds G3BP-1 , 293T cells were cotransfected with expression plasmids encoding HSV-1 ICP8 together with plasmids encoding EGFP-C1 , EGFP-G3BP-wt , -F33W or -F124W . Lysates were immunoprecipitated with anti-GFP and immunoblotted for ICP8 , GFP or actin ( Fig . 7A ) . The results indicate that ICP8 indeed forms a complex with EGFP-G3BP-wt and EGFP-G3BP-F124W , but not with EGFP-C1 or EGFP-G3BP-F33W . These results are strikingly similar to the binding profiles of SFV nsP3 and USP10 ( Fig . 6B and C ) , and strongly suggest that HSV-1 ICP8 forms a complex with G3BP-1 via binding of the FGDF motif at residues 1144–1147 . Since the binding of nsP3 and USP10 to G3BP inhibits SGs , we hypothesized that ICP8 binding might do the same . To investigate this , BHK cells were transfected with an ICP8 expression plasmid , stressed with sodium arsenite and stained for ICP8 as well as for G3BP-1 and TIA-1 to detect SGs . We observed that in 59% of the transfected cells , ICP8 accumulated mainly in the cytoplasm , while in the remaining 41% , ICP8 was localized in the nucleus . In neither case was there a detectable change in the localization of either G3BP-1 or TIA-1 ( S8 Fig . ) , with these proteins exhibiting diffuse cytoplasmic or nuclear staining , respectively . After sodium arsenite stress however , only 10% of cells with predominantly cytoplasmic ICP8 contained TIA-1 and G3BP-1-positive SGs , while 96% of the cells with nuclear ICP8 and a similar proportion of mock-transfected cells had SGs ( Fig . 7B ) . Representative images are provided in S8 Fig . , showing that , when localized in the cytoplasm , ICP8 blocks the induction of SGs after sodium arsenite treatment . Due to the diffuse staining pattern of cytoplasmic ICP8 , it is not possible to discern the proportion that is interacting with G3BP , but the inhibition of SG assembly was profound . Taken together with the results in Fig . 7A , this suggests that ICP8 , like the cellular protein USP10 and SFV nsP3 , interacts with G3BP via its FGDF-motif in a manner which blocks the formation of SGs
This work describes the FGDF motif , shared by proteins from at least two evolutionarily distant viruses and one cellular protein , that mediates strong binding to the multifunctional G3BP proteins . Among its many roles , G3BP is a critical determinant of SG assembly and is one of few proteins whose overexpression induces SGs [2 , 36] . We have shown that the binding of SFV nsP3 , HSV-1 ICP8 and USP10 to G3BP via their FGDF motifs blocks SG formation ( Figs . 2 , 3D and 7B ) . The SG-nucleating function of G3BP requires both the RRM domain and the NTF2-like domain [2] . During stress , the RRM domain likely binds to translationally silent mRNAs or stalled 40S ribosomal subunits , and together with other RNA-binding proteins , targets them for SG inclusion . The SGs are then assembled via protein/protein interactions between SG-critical molecules . The NTF2-like domain , which binds the FGDF motifs , is involved in interactions with another critical regulator of SGs , caprin-1 [7] . It is therefore likely that the FGDF-mediated interactions inhibit or alter the ability of G3BP to form complexes with other proteins during the early stages of SG nucleation , resulting in an inhibition of SG formation . This work therefore reveals a mechanism by which USP10 acts as a negative regulator of SG formation . Furthermore , since G3BP appears to be targeted by many different viruses , its neutralization seems to be important for viral pathogenesis . Our work presents a common mechanism whereby Old World alphaviruses and at least two members of the alphaherpesviruses disrupt SG assembly . Biophysical analyses demonstrated that one 25-mer peptide containing the two FGDF-motifs of SFV nsP3 binds two molecules of G3BP in vitro , suggesting that nsP3 of SFV and the other Old World alphaviruses binds and recruits two molecules of G3BP per nsP3 molecule ( Fig . 4 ) . We have previously shown that early in SFV infection , SGs are quickly disassembled in the vicinity of the newly established viral replication complexes [29] . The stoichiometry of the interaction may explain the rapidity of SG disassembly by nsP3 , despite its relatively low expression compared to G3BP-targeting proteins of other lytic RNA viruses , such as the picornaviruses . It should be noted that a mutated SFV containing only one FGDF motif ( SFV-Δ78 ) exhibited lower levels of G3BP binding and slower replication kinetics compared to wt SFV , but higher levels of G3BP binding and faster replication kinetics compared to SFV-Δ789 , lacking both FGDF motifs [28] . Unlike other proteins with FGDF motifs such as USP10 and ICP8 , the Old World alphavirus nsP3 has evolved to comprise two consecutive FGDF motifs to ensure rapid and efficient SG disassembly . This appears to be the main function of G3BP sequestration by nsP3’s FGDF motifs and represents an efficient evasion strategy , beneficial for virus replication . Herpes simplex viruses block the induction of SGs via multiple mechanisms , highlighting the potent anti-viral effect of SGs . Although early events in viral infection activate PKR , the viral ICP34 . 5 protein promotes the protein phosphatase 1 ( PP1 ) -mediated dephosphorylation of eIF2α and reactivation of translation [37] . HSV-1 mutants lacking the virion host shutoff ( vhs ) protein , an endoribonuclease that degrades cellular and viral mRNA , induce SGs late in infection [38 , 39] . This suggests that vhs may itself have a role in the inhibition of SGs or may alter expression of other SG-modulating viral gene products . HSV-2 also blocks the formation of SGs induced by sodium arsenite but not Pat A [40] . Here , we identify an FGDF motif at the C-termini of the HSV-1 and HSV-2 ICP8 proteins , and demonstrate that HSV-1 ICP8 binds G3BP and blocks SG formation ( Fig . 7 ) . Although ICP8 is a predominantly nuclear protein during HSV-1 infection , a sizeable fraction of the protein remains in the cytoplasm [41] . Functions of that fraction are not well described . Our results suggest that the cytoplasmic fraction of ICP8 inhibits SG assembly or other functions of G3BP . It remains to be determined if ICP8 contributes to the inhibition of SGs during HSV infection . USP10 is a nucleocytoplasmic deubiquitinating enzyme ( DUB ) , originally shown to be regulated by its binding partner G3BP [8] . USP10 deubiquitinates many different proteins of importance in several human diseases and is also a resident SG protein [16 , 22] . Here , we have shown that overexpression of EGFP-USP10 , but not a mutant lacking an intact FGDF-motif , efficiently blocks the formation of SGs ( Fig . 3 ) . Interestingly , recent work has shown that the compound resveratrol can inhibit the interaction of G3BP with USP10 by binding to the NTF2-like domain of G3BP , thereby stimulating USP10 DUB activity and stabilization of p53 [17] . It appears therefore that the FGDF-mediated G3BP/USP10 complex is mutually inhibitory , with G3BP inhibiting the DUB activity of USP10 and USP10 inhibiting the SG nucleating function of G3BP . Elucidation of these inhibitory mechanism ( s ) will require further studies . A molecular model reveals that the FGDF peptide binds tightly into a hydrophobic groove on the surface of the G3BP NTF2-like domain ( Fig . 5 ) . Mutagenesis analyses demonstrated that both phenylalanine residues and the glycine residue are required for binding , with a strong preference for aspartate in the third position . Both phenylalanine side chains fit snugly within the binding site , the glycine is required for flexibility and the aspartate binds to the G3BP residue K123 . Both phenylalanine and the glycine residues are fully conserved in most Old World alphavirus nsP3 sequences and in the USP10 proteins of all higher eukaryotes . The aspartate residue in the third position of the motif is conserved in many of the Old World alphavirus nsP3 sequences except that of Sindbis virus , in which it is a serine . A serine is also found in this position in the Arabidopsis thaliana USP10 gene , while this residue is a glutamate in avian and fish USP10 genes . We note the congruence of our biochemical , phylogenetic and structural analyses and propose that the core binding motif consists of FGxF , in which x can be aspartate , glutamate or serine . While mutation of the aspartate ( D7 ) immediately downstream of the FGDF motif in SFV nsP3 had little effect on binding ( Fig . 1B ) , it is notable that the FGDF motifs in SFV nsP3 , USP10 and HSV-1 ICP8 are all followed by at least two acidic residues within the downstream four residues . In our molecular model , we observe that these residues are likely involved in interactions with the basic residues localized the N-terminus of G3BP , which is required for FGDF binding ( Fig . 6E ) and we propose that they constitute an important part of the motif that further stabilizes the complex . Using these criteria , we present a list of human and viral proteins that contain this motif , and therefore are candidate G3BP-binding proteins ( S1 and S2 Tables ) . In conclusion , our work describes a motif shared by three otherwise very different proteins , and potentially others , that mediates binding to G3BP and thereby inhibits SG formation . Our three-dimensional model provides a structural understanding of the G3BP/FGDF interaction and will form the basis for the design of pharmaceuticals to target this interaction with a therapeutic potential for a range of viral infections as well as cancers .
Expression vectors for EGFP-nsP3-31 [28] , nsP3-BAP [27 , 42] , G3BP-NTF2 [27] , and HSV-1 ICP8 [43] were described previously . pEGFP-USP101–40 wt sequences and corresponding alanine mutants were obtained from GeneArt , and ligated between the BglII and EcoRI sites of pEGFP-C1 . Construction of the infectious clone pCMV-SFV-F3A: The PCR product derived from primers 1 and 2 ( S3 Table ) and the PCR product derived from primers 3 and 4 were fused by a one-step PCR in a molar ratio of 1:1 . The DNA was denatured and annealed at 46˚C for 2 min . These partially double-stranded molecules were made fully double stranded by extension at 72˚C for 3 min . The fusion DNA was then amplified by using primer 1 and 4 for 25 cycles of PCR consisting of treatment at 95˚C for 30 s , 69˚C for 30 s , and 72˚C for 2 min , followed by a final extension at 72˚C for 5 min . The derived PCR product was purified and subcloned into pTZ57R/T plasmid ( Thermo Scientific ) . The resulting pTZ57R/T-F3A plasmid was digested with XhoI and BglII and religated to the similar digested pCMV-SFV4 vector [44] . The presence of mutations was confirmed by sequencing . Construction of pEGFP-G3BP-F33W and -F124W: The PCR product derived from primers 5 and 6 ( F33W ) or primers 5 and 7 ( F124W ) and the product derived from primers 8 and 10 ( F33W ) or primers 9 and 10 ( F124W ) were fused by a one-step PCR in a molar ratio of 1:1 . The DNA was denatured and annealed at 33˚C for 2 min . These partially double-stranded molecules were made fully double stranded by extension at 72˚C for 3 min . The fusion DNAs containing the F33W mutation or F124W mutation were then amplified by using primers 5 and 10 for 25 cycles of PCR consisting of treatment at 95˚C for 30 s , 56˚C for 30 s , and 72˚C for 2 min , followed by a final extension at 72˚C for 5 min . The derived PCR product was purified and subcloned into pTZ57R/T plasmid ( Thermo Scientific ) . The resulting pTZ57R/T-G3BP1-F33W , -F124W plasmid was digested with BglII and EcoRI and religated to the similar digested pEGFP-C1-G3BP1 vector . The presence of mutations was confirmed by sequencing . All cell lines were maintained as previously described [28 , 45 , 46] . Where indicated , cells were stressed by addition of sodium arsenite ( 0 . 5 mM ) or pateamine A ( 100 nM ) in complete medium for 60 min . Cells were transfected with Lipofectamine 2000 ( Invitrogen ) reagent according to the manufacturer’s instructions . Virus titration was performed by plaque assay , as previously described [28] Immunofluorescence , immunoprecipitations and immunoblotting were performed as described previously ( Panas et al . , 2012 ) . For details of all antibodies used , see S4 Table . His-tagged G3BP-NTF2 was expressed in E . coli BL21 T7 Express cells and purified using HisTrap columns ( GE Healthcare ) . Before ITC , G3BP-NTF2 was eluted from a Superdex 75 HiLoad 16/60 ( GE Healthcare ) SEC column equilibrated in ITC-buffer ( 25 mM HEPES , 150 mM NaCl , 10 mM MgCl2 , 10% glycerol , pH 7 . 5 ) at a retention volume of 60 mL . The peptides nsP3-25 and nsP3-25-mut were dissolved in ITC buffer at a concentration of 500 μM and dialyzed extensively against the same buffer . ITC measurements were performed using an ITC200 titration calorimeter ( GE Healthcare ) . The cell temperature was set to 37°C , the reference power to 7 μCal/sec and the syringe stirring speed to 1000 rpm . G3BP-NTF2 was loaded into the cell . The peptides were titrated in 48 injections , each injection with a volume of 750 nL , a duration time of 1 . 5 sec and a waiting time between the injections of 150 sec . The first injection was performed using a volume of 300 nL , a duration time of 0 . 6 sec and a spacing time of 120 sec . Background measurements were performed with buffer injected into the protein solution , and peptide into the buffer solution . Data were analyzed using the Origin software as included in the instrument package . Before analysis , 40 μM G3BP-NTF2 and 400 μM nsP3-25 were mixed and incubated for 1h in SEC-buffer ( 25 mM HEPES , 300 mM NaCl , 5 mM MgCl2 , 10% glycerol , pH 7 ) . A sample volume of 100 μL was injected onto the Superdex 75 10/300 GL column equilibrated in SEC-buffer . G3BP-NTF2 alone was used a control . The apparent molecular weights of the eluted proteins were calculated from their retention volumes as described in the gel filtration LMW/HMW calibration kits assuming similar globular shapes for the analyzed proteins and calibration standards ( GE Healthcare ) . . | Stress granules ( SGs ) are dynamic aggregates of proteins and translationally silenced mRNA that are formed in cells upon various stress conditions , such as virus infection . SGs are thought to be antiviral , and many viruses have hence evolved countermeasures to prevent their formation , often targeting the essential SG protein G3BP . Here , we show that several otherwise unrelated viral and cellular proteins all bind G3BP with the sequence motif FGDF , and thereby repress SG formation: the non-structural protein 3 ( nsP3 ) of the Old World alphavirus Semliki Forest virus ( a close relative of the emerging , highly pathogenic Chikungunya virus ) ; the protein ICP8 of herpes simplex virus; and in addition , the cellular protein USP10 ( an SG component and protein deubiquitinase that stabilises e . g . the tumor suppressor p53 ) . In this work , we also present and validate a model of the three-dimensional structure of G3BP bound to an FGDF-containing peptide . The FGDF-mediated G3BP binding represents an attractive target for therapeutic interventions against a range of diverse viral infections , and may also regulate the p53-stabilising function of USP10 in cancers . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2015 | Viral and Cellular Proteins Containing FGDF Motifs Bind G3BP to Block Stress Granule Formation |
The sense of taste is fundamental to our ability to ingest nutritious substances and to detect and avoid potentially toxic ones . Sensory taste buds are housed in papillae that develop from epithelial placodes . Three distinct types of gustatory papillae reside on the rodent tongue: small fungiform papillae are found in the anterior tongue , whereas the posterior tongue contains the larger foliate papillae and a single midline circumvallate papilla ( CVP ) . Despite the great variation in the number of CVPs in mammals , its importance in taste function , and its status as the largest of the taste papillae , very little is known about the development of this structure . Here , we report that a balance between Sprouty ( Spry ) genes and Fgf10 , which respectively antagonize and activate receptor tyrosine kinase ( RTK ) signaling , regulates the number of CVPs . Deletion of Spry2 alone resulted in duplication of the CVP as a result of an increase in the size of the placode progenitor field , and Spry1−/−;Spry2−/− embryos had multiple CVPs , demonstrating the redundancy of Sprouty genes in regulating the progenitor field size . By contrast , deletion of Fgf10 led to absence of the CVP , identifying FGF10 as the first inductive , mesenchyme-derived factor for taste papillae . Our results provide the first demonstration of the role of epithelial-mesenchymal FGF signaling in taste papilla development , indicate that regulation of the progenitor field size by FGF signaling is a critical determinant of papilla number , and suggest that the great variation in CVP number among mammalian species may be linked to levels of signaling by the FGF pathway .
Taste sensory capability is mediated by aggregates of receptor cells , called taste buds , which reside within the oral and pharyngeal cavities . The majority of taste buds in mammals reside on the tongue in epithelial-mesenchymal specializations termed gustatory papillae . In the rodent tongue , the smaller fungiform papillae , each of which possesses a single taste bud , are found in a distributed array on the anterior tongue . By contrast , the larger bilateral foliate papillae and a single midline circumvallate papilla ( CVP ) each house hundreds of buds and reside on the posterior tongue ( Figure 1A ) . Recently , there has been increasing recognition that anterior fungiform taste buds differ from those of the posterior CVP in terms of both gene expression and taste function [1]–[4] . In recent years , significant progress has been made in defining the molecular regulation of fungiform development . Fungiform papillae initially form as placodes that subsequently undergo epithelial morphogenesis and acquire a mesenchymal core [5] in a process that is similar to morphogenesis of other vertebrate epithelial specializations , such as hair , teeth , and mammary glands [6] , [7] . The development of these other organs requires signaling between epithelium and mesenchyme , suggesting that such epithelial-mesenchymal interactions are also involved in patterning and morphogenesis of taste placodes . However , expression of all of the key signaling factors implicated in taste placode development – including Sonic Hedgehog ( SHH ) , Bone Morphogenetic Proteins ( BMPs ) , Epidermal Growth Factor ( EGF ) , and WNTs – is restricted to the epithelium [8]–[15] . Thus , to date , no inductive , mesenchyme-derived factor involved in taste development has been identified . Despite its importance in taste function and its status as the largest of the taste papillae , very little is known about the genes involved in the development of the CVP . Like fungiform placodes , the CVP forms as an initial epithelial thickening that undergoes complex morphogenesis to form a large papilla . However , it appears that genes known to regulate fungiform development do not function similarly in development of the CVP . For example , inhibition of SHH results in more and larger fungiform placodes , but has no effect on the CVP [11] . In addition , BMP7 and its antagonist follistatin have significant functions in fungiform development , but the CVP appears unaffected by inactivation of either gene [16] . These differences may be ascribed to the distinct embryonic origins of the anterior tongue , which is thought to be derived from ectoderm , whereas the posterior tongue likely has endodermal origins [4] . Expression of several Fibroblast Growth Factors ( FGFs ) and their receptors has previously been detected in the developing tongue [17] . Therefore , we hypothesized that Sprouty ( Spry ) genes , which antagonize several receptor-tyrosine kinase ( RTK ) signaling pathways including those triggered by FGFs [18]–[20] , may play a role in the development of taste papillae . Originally , spry was identified as a regulator of tracheal branching in Drosophila [21] , and it was later found that three of the four mouse Sprouty genes ( Spry1 , Spry2 , and Spry4 ) are expressed during embryonic development [22] , [23] . Here , we used mouse genetic models to show that FGF signaling is required for CVP formation . We found that Spry1 and Spry2 antagonize signaling by FGF10 to restrict the size of the progenitor field of the circumvallate ( CV ) placode , such that loss of Sprouty function results in a dramatic expansion of the CV placode as it first forms . In adult Spry2 mutants , a striking and complete duplication of the CVP emerges , whereas embryos lacking both Spry1 and Spry2 have multiple CVPs . Our findings thus represent the first example of molecular genetic regulation of taste organs in the posterior tongue . We found that Fgf10 , which is expressed exclusively in the mesenchyme underlying the CV placode , is required for formation of the CVP , and thus we provide the first example of an inductive , mesenchymal signal involved in specifying the epithelial domain of a developing taste papilla . Further , while FGF10 signaling drives CVP development and Spry1 and Spry2 repress this process , fungiform taste papillae are oppositely affected by the loss of these genes: Fgf10−/− tongues appear to possess more and larger fungiform papillae , and the loss of Sprouty genes results in fewer fungiform papillae . Thus , these results demonstrate that molecular mechanisms regulating development of anterior and posterior taste organs differ considerably . Finally , we postulate that the role of FGF signaling in defining the size of the CV progenitor field in mice may underlie the large variation in CVP number across mammalian taxa , and that changes in FGF signaling during evolution may have caused expansion of the initial progenitor field to allow formation of multiple CVPs in some species .
Wild-type mice possess a single CVP in the midline of the posterior tongue; foliate papillae reside on the lateral tongue and multiple fungiform papillae populate the anterior tongue ( Figure 1A , 1B ) . In Spry2−/− mice , we observed a duplication of the CVP in the anterior-posterior orientation at embryonic day ( E ) 14 . 5 using Shh expression as a marker ( Figure 1C ) . By contrast , fungiform placodes in the anterior tongue , which also express Shh , were significantly reduced in number ( 61±4 . 37 in wild-type versus 33±2 . 02 in Spry2−/− embryos; p = 0 . 0002; n = 6 embryos per genotype ) . Because alterations in fungiform papilla development had been reported in other mutants [9]–[15] , we chose to pursue the novel CVP duplication phenotype . The presence of two CVPs in the embryonic tongue was confirmed using scanning electron microscopy ( SEM; Figure 1D-1E' ) . To test whether the duplicated CVP persisted into adulthood , we used DAPI staining in adult tongues and again identified two discrete papillae ( Figure 1F-1G' ) . Notably , both CVPs in Spry2−/− mice house fully differentiated taste buds containing the three types of taste receptor cells , and both showed positive staining for b3-tubulin , indicating that taste buds and the papillae are innervated ( Figure S1 ) . Taken together , these data indicate that the duplication in Spry2−/− embryos leads to two functional adult CVPs . To reconstruct the development of the CVP , closely staged specimens between E11 . 5 to E14 . 5 were stained with E-cadherin and imaged in whole-mount ( Figure 2A-2N ) , as well as analyzed by H&E staining of sections ( Figure 2O-2X ) . In wild-type embryos , CVP development was initiated shortly after the tongue rudiment formed by fusion of the lateral lingual swellings , and it was detected as a more compact collection of epithelial cells at E11 . 5 ( Figure 2A ) [24] . At E12 . 5 , a placodal condensation was readily observable on coronal sections ( Figure 2O ) . Between E13 and E13 . 5 , the wild-type placode underwent major morphological changes as the epithelial trenches characteristic of the CVP began to form ( Figure 2P , 2Q ) . The growth of the trenches was coincident with the initiation of innervation at approximately E13 ( Figure S2 ) , and the trenches continued to grow into the mesenchyme at E14 . 5 ( Figure 2G , 2S ) [24] , [25] . In contrast to the wild-type , the CV placode in Spry2−/− embryos was larger at E11 . 5 ( compare Figure 2A , 2H ) . Relative to the wild-type , the mutant placode continued to grow larger both laterally and along the anterior-posterior axis between E11 . 5 and E13 . 5 ( Figure 2B-2E , 2I-2L ) . By E14 , two distinct structures were observed in Spry2−/− embryos , with the anterior of the two papillae usually appearing smaller than the posterior at E14 and E14 . 5 ( Figure 2M , 2N ) . In addition , CVPs in Spry2−/− embryos showed a raised dome shape compared to the flatter shape of the wild-type CVP ( Figure 2R , 2S , 2W , 2X ) . Several possible mechanisms could account for the duplication of the CVP at ∼E14 , including a timer that causes the organ to split at a certain point after its development starts or a threshold that causes splitting after a critical size limit is achieved . To address this issue , we quantified the size of the CV placode and CVP in wild-type and Spry2−/− tongues between E13 to E14 , corresponding to the time points before , during and after CVP duplication ( Figure S2J ) . Relative to wild-type mice , the mutant CV placode was already significantly larger at E13 and E13 . 5 , but when CVP duplication was observed at E14 , there was a further and dramatic increase in the total area of the papillae . Thus , our data suggest that there is a critical threshold size for the CV placode , beyond which the placode destabilizes , leading to CVP duplication . Sox2 is one of the earliest markers of taste placodes and is required for taste bud development [26] . Sox2 expression was expanded in the placode domain in both anterior-posterior and lateral directions in Spry2−/− embryos compared to wild-type littermates at both E11 . 5 ( data not shown ) and E12 . 5 ( Figure 3A-3B' ) . Next , we analyzed the expression of three additional factors that mark the early placode: Shh , Bmp7 , and Wnt10b ( Figure 3C-3H' ) . Shh has previously been shown to be expressed in the developing CVP [9] , [27] , and in fungiform placodes , Shh is expressed specifically by taste bud progenitors [28] . Bmp7 and Wnt10b have also been implicated in development of fungiform papillae [15] , [16] . The expression domains of all these markers were expanded in the CV placode of Spry2−/− embryos relative to wild-type ( Figure 3C-3H' ) . We quantified the differences in expression levels by qPCR , and interestingly , whereas Sox2 and Bmp7 levels were strongly upregulated in the mutants , consistent with the expanded expression domains seen in whole mount , Shh and Wnt10b expression levels were not dramatically different in the mutants ( Figure S3A ) . This is consistent with the decreased intensity of Shh and Wnt10b staining despite an increased CV placodal domain size ( Figure 3D , 3D' , 3H , 3H' ) . We next asked whether there was a difference in the number of cells in the placode between wild-type and mutant . We quantified the number of cells in the CV placode of Spry2−/− embryos relative to wild-type littermates at E12 . 5 from three-dimensional confocal images of E-cadherin-positive stained cells ( Figure 3I-3K ) . We counted cells in 10 mm intervals beginning at the deepest tip of invaginated cells using ventral views; the borders of the placode were defined by the invagination . This analysis showed that already by the mid-placode stage , the mutant placode had almost twice as many cells as the wild-type . Because differences in proliferation or cell death could account for the larger placode and the morphogenetic abnormalities in the CVPs of Spry2 mutants , we assayed proliferation by PCNA staining ( Figure 3L-3N ) , and cell death by TUNEL staining ( Figure S3B-S3D ) . We detected no differences between wild-type and mutant CV placode at E12 . 5 for either assay , suggesting that a larger number of cells was recruited into the placode at the earliest stages of placodogenesis . Thus , our data indicate that inactivation of Spry2 leads to a larger progenitor field , as detected by multiple molecular markers , such that the duplication in the Spry2 mutant results from recruitment of more cells into the CV placode at the earliest stages of its development . Because Sprouty genes are often co-expressed during development [23] , [29] , we assayed for expression of Spry1 , Spry2 , and Spry4 in CV placodes at E12 . 5 ( Figure 4A-4C ) ; Spry3 was not analyzed due to its lack of expression in embryonic craniofacial tissues [23] . We found that Spry1 and Spry2 were both expressed in the epithelium of the CV placode at E12 . 5 , whereas no expression was observed for Spry4 ( Figure 4A-4C ) . The expression of Etv5 ( Erm ) , which is a target of FGF signaling [30] , was upregulated in Spry2−/− embryos ( compare Figure 4D , 4E ) , consistent with increased FGF signaling observed with Sprouty loss-of-function in other developmental contexts [31]–[34] . To quantify expression levels , RNA was extracted from the area containing the CV placode at E12 . 5 and analyzed by qPCR , and a three-fold increase in Etv5 was observed ( Figure 4F ) . As expected , essentially no Spry2 expression was detected in the Spry2 nulls , but interestingly , Spry1 expression was dramatically increased in Spry2−/− mice ( Figure 4F ) . This upregulation is consistent with the known role of Sprouty genes as FGF targets [21] , [23] , and it suggested that increased expression of Spry1 may serve a compensatory role in Spry2−/− mice . qPCR showed that Spry4 was expressed at very low levels , with no detectable difference between wild-type and Spry2−/− littermates ( data not shown ) . The epithelial expression of Spry1 , Spry2 , and Etv5 , all of which are expressed in response to RTK activity [21] , [23] , [30] , strongly suggested the presence of a mesenchymal RTK ligand that signals to the epithelium . To identify such a ligand , we surveyed the expression of a number of candidate RTKs and their cognate ligands by ISH , including Fgfr1-4 , Egfr , Fgf7 , Fgf8 , Fgf9 , and Fgf10 ( data not shown ) . Of the ligands we examined , only Fgf10 was expressed in the mesenchyme subjacent to the epithelium along the midline of the posterior tongue at E11 . 5 and E12 . 5 ( Figure 5A-5B' ) . Two FGF receptors , Fgfr2 and Fgfr3 , were also expressed in the CV placode , whereas no expression of Fgfr1 or Fgfr4 was detected ( Figure 5C-5F ) . To further test the role of FGF signaling in CVP development , we isolated tongues from wild-type embryos at E11 . 5 and E12 . 5 and grew them in vitro either in the absence or presence of SU5402 [35] , a potent FGF receptor inhibitor ( Figure S4 ) . In the absence of SU5402 , we found the formation of a single CVP . The inhibition of FGF signaling by SU5402 led to the absence of the CVP in tongues cultured at E11 . 5 and to either an absent or a malformed CVP in tongues cultured at E12 . 5 . These observations demonstrate the requirement of FGF signaling during CVP formation at E11 . 5 and to a lesser extent at E12 . 5 . Because Fgf10 was expressed directly underneath the developing CV placode , we hypothesized that the ligand encoded by this gene may play a role in CV development , and therefore we examined the CVP in Fgf10−/− embryos . Strikingly , the deletion of Fgf10 led to the complete absence of the CVP ( Figure 5H-5H” ) , indicating that this gene plays a critical inductive role in development of this structure . Next , we reasoned that the combined deletion of Fgf10 and Spry2 may balance each other and rescue CVP development , and indeed , we found a single CVP in Fgf10−/−;Spry2−/− mice ( Figure 5I-5I” ) . This result also implies the presence of a yet-unknown secondary ligand that can fulfill the function of FGF10 if the epithelium is rendered hypersensitive to RTK signaling by inactivation of Spry2 . Taken together , our data indicate that the FGF signaling pathway is critical for proper CVP development and that it determines the final number of CVPs . Based on the overlapping expression of Spry1 and Spry2 in the CV placode and the upregulation of Spry1 in Spry2−/− embryos ( Figure 4F ) , we hypothesized that these two family members may be partially redundant . We therefore generated and analyzed Spry1−/− as well as Spry1−/−;Spry2−/− ( DKO ) embryos . No difference in the number or morphology of the CVP was detected in Spry1−/− embryos ( data not shown ) , indicating that Spry1 is not required for CVP development if Spry2 is present . In contrast , CVPs from DKO embryos were dramatically abnormal relative to control Spry1+/−;Spry2+/− ( DHet ) littermates , whose tongues were indistinguishable from those of wild-types both as embryos and as adults ( data not shown ) . At E12 . 5 , the expression domains of Etv5 and Sox2 were expanded in the CV placodes of DKO mice relative to DHet littermates ( Figure 6A , 6B , 6F , 6G ) . The expanded expression domains in the DKOs were greater than those seen in Spry2−/− mice ( compare to Figure 4D , 4E and Figure 3A-3B' ) . At later stages , in contrast to the single duplication seen in Spry2−/− mice , the DKO embryos had multiple CVPs ( Figure 6C , 6E , 6H , 6J ) . The multiple CVPs in DKO embryos were smaller than the single CVP in controls and showed pronounced raised domes ( Figure 6D , 6I ) similar to Spry2−/− embryos ( Figure 2S , 2X ) . Immunofluorescence staining of β3-tubulin revealed innervation of the multiple CVPs ( Figure S5 ) .
Antagonism of FGF signaling by Sprouty genes occurs during development of a number of organs [31]–[34] . Because FGFs and their receptors had previously been detected in the developing tongue [17] , we hypothesized that Sprouty genes , which modulate several RTK signaling pathways including those triggered by FGFs [18]–[20] , may play a role in taste papillae development . Whereas deletion of Spry2 led to a duplication of the CVP , Fgf10−/− mice showed a complete loss of the CVP , and importantly , CVP development was rescued in compound Spry2−/−; Fgf10−/− mutants . These results confirm the specific antagonism of FGF10 signaling by SPRY2 and establish the importance of the FGF signaling pathway in determining CVP number . Additionally , the results with the Spry2−/−;Fgf10−/− mutants indicate the presence of an unknown , secondary factor; such a factor could induce CVP formation in lieu of FGF10 when the epithelium is hypersensitive to RTK signaling due to the absence of Spry2 . Because of the overlapping expression profiles of Spry1 and Spry2 , and the upregulation of Spry1 in the Spry2−/− CV placode , we generated Spry1−/−; Spry2−/− embryos . We observed multiple CVPs along the A-P axis in the compound mutant embryos , demonstrating the redundancy of Sprouty genes , for which there is precedent in other tissues [32] , in regulating CVP development . Although a slightly dysplastic CVP was recently reported in ectodysplasin mutant mice [36] , the phenotypes in Fgf10−/− , Spry2−/− , and Spry1−/−;Spry2−/− mice are the first genetic abnormalities reported involving the regulation of CVP number . Despite indications that mesenchyme-derived factors are involved in lingual papillae development [5] , [37] , the only mesenchymal factor implicated in fungiform papillae formation to date is follistatin , an inhibitor of BMP signaling [16] . Therefore , FGF10 is the first inductive , mesenchyme-derived factor to be identified in taste papilla development . The duplication of the CVP at E14 in Spry2−/− embryos was preceded by an increase in placode size . The CV placode in mutant mice was significantly larger at E13 and E13 . 5 relative to wild-type , and at E14 there was a dramatic increase in total CVP size as duplication occurred ( Figure S2J ) . Thus , there appears to be a critical size threshold for the CV placode , beyond which the placode destabilizes , resulting in splitting and duplication . Recently , Munne et al . [38] reported that , during incisor tooth development , the impairment of BMP4 signaling or an increase in Activin concentration led to the destabilization of the large , single placode and the formation of multiple incisors . Future experiments will enable comparison of the relative roles of different signaling pathways in maintenance of epithelial placode integrity . The increase in placode size at E13 and E13 . 5 was preceded at E11 . 5 and E12 . 5 by the expansion of the expression domains for various genes , such as Etv5 , Sox2 , Shh , Bmp7 , and Wnt10b . By E12 . 5 , the mutant CV placode already had more cells than the control , and we did not detect any apoptosis in the CV placode or any differences in the percentage of proliferating cells . Thus , the larger CV placode in the Spry2 mutant appears to result at least in part from an increase in the recruitment of CV placodal cells . At the earliest stages of placode formation , this could potentially occur either through specification of more placodal precursors or through migration of more cells into the placodal domain . Future studies will be needed to distinguish between these possibilities , but interestingly , several reports have pointed to a role for Sprouty genes in cell migration [39]–[41] . In the CV placode of Spry1−/−;Spry2−/− tongues , there were further increases in the expression domains of Etv5 and Sox2 beyond what was seen in the Spry2−/− alone , demonstrating the redundancy of Spry1 and Spry2 in this organ ( Figure 6A , 6B , 6F , 6G ) . The compound deletion of Spry1 and Spry2 led to an even larger placode ( Figure 6A , 6B , 6F , 6G ) , which resulted in more than two CVPs . Whereas the deletion of Spry2 led to CVP duplication , it also resulted in a marked decrease in the number of fungiform papillae ( Figure 1B , 1C ) . In contrast , in Fgf10−/−mice , there was an absence of the CVP , whereas the fungiform papillae appeared to be larger and more closely spaced ( Figure 4G , 4H ) . Thus , Fgf10 and Sprouty genes differentially affect the anterior versus posterior taste fields . Together with the previous studies showing effects of SHH and BMP7 on anterior but not posterior taste fields , these results provide further evidence for important developmental differences between these fields [9] , [10] , [12] , [16] . This observation is consistent with the notion that fungiform papillae in the anterior tongue are derived from ectoderm , whereas the CVP is likely derived from endoderm [4] . The diverse responses of fungiform versus CV papillae to developmental factors have only recently been appreciated and will require further efforts to tease apart . Although most rodents , including the mouse , possess a single midline CVP in the posterior tongue , there is great variation in mammalian CVP number ( Figure S6B , Figure S7 ) . For example , humans possess anywhere from 3 to 14 CVPs in an inverted V or Y orientation [42] . Other mammals such as the hyrax [43] , [44] and hippopotamus [45] possess no CVP , whereas siamang , chimpanzee , gorilla [46] and lemur [47] possess at least two CVPs along the midline , in addition to varying numbers of lateral CVPs ( Figure S6A ) . We have shown that modifications in FGF signaling can lead to increased or decreased numbers of CVPs . Thus , we speculate that changes in levels of signaling through this pathway provide an attractive candidate for producing variation in CVP number , and in particular , for generating CVPs that are multiplied along the midline or anterior-posterior axis . It is interesting that both Fgf10 and Etv5 , an indicator of FGF signaling , were expressed along the midline , which correlates with anterior-posterior CVP multiplication in Spry2−/− and Spry1−/−;Spry2−/− mice . The effect of CVP number on taste is currently unclear . Correlation between taste sensitivity and the number of taste buds within the CVP has been previously reported [48] . However , because there is variation in the number of taste buds per CVP , as well as in the number of taste cells within each taste bud [49] , there is no clear indication of what the number of CVPs reveals about taste preferences . Whether the variation in mammalian CVP number provides evolutionary advantages in terms of identification of nutritious substances and detection and avoidance of potentially toxic ones remains to be elucidated . In 1950 , Spuhler [42] postulated that the variation in CVP number observed in humans was likely due to genetic factors involving at least 5 multiple alleles . Our studies indicate , for the first time , that the perturbation of a single gene such as Fgf10 or Spry2 may be sufficient to confer the vast genetic variation in mammalian CVP number .
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 UCSF Institutional Animal Care and Use Committee ( Protocol Number: AN084146 ) . All efforts were made to minimize animal suffering . Mouse lines carrying null or floxed alleles of Spry1 [31] , Spry2 [34] , and Fgf10 [50] or β-actin Cre transgenes [51] were maintained and genotyped as described . Spry1−/−;Spry2−/− mutant embryos were generated by crossing β-actin Cre;Spry1+/−;Spry2+/− males with Spry1flox/flox;Spry2flox/flox females ( double mutants produced at an expected Mendelian frequency of 1∶4 ) . Tongues and taste papillae from double heterozygous embryos ( Spry1+/−;Spry2+/− ) were indistinguishable from wild-type CD-1 embryos and served as controls . Mice were mated overnight , and the presence of a vaginal plug indicated embryonic day ( E ) 0 . 5 . Embryonic and adult tongues were fixed overnight in 4% paraformaldehyde at 4°C . For sections , tongues were dehydrated , embedded in paraffin wax , and serially sectioned at 7 µm . Histological sections were stained with haematoxylin and eosin ( H&E ) . Whole-mount ISH using digoxigenin-labeled RNA probes was performed according to standard protocols . RNA probes were generated from plasmids containing fragments of Shh , Spry1 , Spry2 , Spry4 , Fgf10 , and Sox2 or from PCR-amplified fragments of Wnt10b and Bmp7 . A 10 minute 2% H2O2 treatment was performed on tissues E14 . 5 and older . Stained specimens were incubated overnight in 30% sucrose , embedded in tissue freezing media ( Triangle Biomedical Sciences , Durham , N . C . ) , and cryosectioned using a Leica CM1900 at 18 µm intervals . Whole-mount IHC was performed according to a published protocol [52] . Rat anti-E-cadherin ( 1∶1000; Invitrogen Cat . 1300 ) or mouse anti-β3 tubulin ( 1∶250; R&D Systems ) were applied followed by incubation in goat anti-rat AlexaFluor 555 secondary antibody ( 1∶250 , Invitrogen ) . For β3-tubulin staining , the Mouse on Mouse kit ( Vectastain ) was used . Specimens were counterstained with DAPI . For IHC on sectioned embryonic specimens the same procedure was used after a rehydration step . For IHC on adult tongues , tissue was processed according to a published protocol [28] for markers of three types of taste receptor cells: type 1 , rabbit anti-NTPdase2 ( Nucleoside triphosphate diphosphohydrolase-2 , 1∶1000 ) is the ecto-ATPase of type I cells in taste buds [53]; type II , monoclonal anti-IP3R3 ( Receptor for inositol 1 , 4 , 5-trisphosphate , 1∶1000; BD Transduction ) is a second messenger that mediates the release of intracellular calcium [54]; and type III , rabbit anti-NCAM ( Neural cell adhesion molecule , 1∶1000 ) [55] . Secondary antibody was goat anti-rabbit Alexa-488 ( 1∶500; Invitrogen ) ; this was applied for 2 hours and counterstained with propidium iodide . For quantification of placode size , wild-type and Spry2−/− tongues between E13 and E14 were stained with anti-E-Cadherin antibody and the area measured using ImageJ software . Apoptosis in the CV placode at E12 . 5 was measured using the In situ Cell Death Detection kit ( Roche ) following manufacturer's protocol . Proliferating cells were identified by anti-PCNA immunofluorescence staining or injection of 1 mg BrdU for 2 hours followed by staining with anti-BrdU antibody ( Invitrogen ) . Anti-PCNA stained and total ( i . e . DAPI-stained ) number of cells were counted using ImageJ software and presented as a percentage of proliferating cells . The total number of cells in the CV placode was quantified from confocal images of E-cadherin stained sections using Volocity5 software ( Improvision ) . qPCR reactions were performed using the GoTaq qPCR Master Mix ( Promega ) in a Mastercycler Realplex ( Eppendorf ) . All primers were designed using PerlPrimer3 software [56]; sequences are available upon request . qPCR conditions were as follows: 95°C , 2 minutes; 40 cycles at 95°C , 15 seconds; 58°C , 15 seconds; 68°C , 20 seconds; followed by a melting curve gradient . Expression levels for the genes of interest were normalized to levels of L19 and are presented as levels relative to wild-type . Developing tongues at E11 . 5 and E12 . 5 were isolated and cultured on 0 . 45 µm Millicell-HA membranes ( Millipore ) in F12/DMEM medium ( GIBCO/Invitrogen ) containing 1% FBS , 2% B27 culture supplement ( GIBCO/Invitrogen ) , and antibiotics . FGF signaling was inhibited by addition of 25 µM SU5402 suspended in DMSO ( Calbiochem ) , and an equivalent volume of DMSO was added to control wells . After 3 days in culture , the tongues were fixed in 4% paraformaldehyde for 2 h and analyzed . Fluorescent and bright field images were taken using a Leica DM5000B with a Leica DFC500 camera . For confocal images , a Leica SP5 Confocal was used . Unless otherwise noted , all experiments were performed independently in triplicate on at least three different specimens ( N≥3 ) , and when applicable , presented as an average ± standard deviation . Unpaired Student t-test was used to determine p-values and p<0 . 01 was deemed to be significant . The CVP phenotype observed in Spry2 null mice was 100% penetrant ( n>12 ) ; the loss of CVP in Fgf10 null mice was observed in 100% of the mice ( n = 6 ) , however , there were indications of small trenches or invaginations ( although absence of innervation ) in 66% of the embryos; the rescue of the single CVP in the double Fgf10;Spry2 null mice was 100% penetrant ( n = 4 ) ; the presence of multiple ( i . e . ≥3 ) CVPs in Spry1;Spry2 null mice was 86% penetrant ( n = 6 ) . The Fisher exact probability test was used to determine the p-values of the tongue culture experiments . | The sense of taste is important for an animal's ability to survive and thrive , because it enables discrimination between nutritious substances and toxins . Taste buds are housed largely on the tongue in structures called papillae; of the three types of gustatory papillae , the circumvallate papilla ( CVP ) is the largest . In rodents , a single CVP is located in the posterior midline of the tongue housing hundreds of taste buds , whereas in other mammals up to dozens of CVPs can be found . However , despite the great variation in the number of CVPs in mammals , its status as the largest of the taste papillae , and its importance in taste function , very little is known about its development . We identified members of the FGF signaling pathway as determinants of CVP number . We propose that perturbations to the FGF signaling pathway may have been involved in the dramatic differences in CVP number that arose during mammalian evolution . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"developmental",
"biology",
"biology",
"evolutionary",
"biology"
] | 2011 | FGF Signaling Regulates the Number of Posterior Taste Papillae by Controlling Progenitor Field Size |
Ciliopathies are a group of genetic multi-systemic disorders related to dysfunction of the primary cilium , a sensory organelle present at the cell surface that regulates key signaling pathways during development and tissue homeostasis . In order to identify novel genes whose mutations would cause severe developmental ciliopathies , >500 patients/fetuses were analyzed by a targeted high throughput sequencing approach allowing exome sequencing of >1200 ciliary genes . NEK8/NPHP9 mutations were identified in five cases with severe overlapping phenotypes including renal cystic dysplasia/hypodysplasia , situs inversus , cardiopathy with hypertrophic septum and bile duct paucity . These cases highlight a genotype-phenotype correlation , with missense and nonsense mutations associated with hypodysplasia and enlarged cystic organs , respectively . Functional analyses of NEK8 mutations in patient fibroblasts and mIMCD3 cells showed that these mutations differentially affect ciliogenesis , proliferation/apoptosis/DNA damage response , as well as epithelial morphogenesis . Notably , missense mutations exacerbated some of the defects due to NEK8 loss of function , highlighting their likely gain-of-function effect . We also showed that NEK8 missense and loss-of-function mutations differentially affect the regulation of the main Hippo signaling effector , YAP , as well as the expression of its target genes in patient fibroblasts and renal cells . YAP imbalance was also observed in enlarged spheroids of Nek8-invalidated renal epithelial cells grown in 3D culture , as well as in cystic kidneys of Jck mice . Moreover , co-injection of nek8 MO with WT or mutated NEK8-GFP RNA in zebrafish embryos led to shortened dorsally curved body axis , similar to embryos injected with human YAP RNA . Finally , treatment with Verteporfin , an inhibitor of YAP transcriptional activity , partially rescued the 3D spheroid defects of Nek8-invalidated cells and the abnormalities of NEK8-overexpressing zebrafish embryos . Altogether , our study demonstrates that NEK8 human mutations cause major organ developmental defects due to altered ciliogenesis and cell differentiation/proliferation through deregulation of the Hippo pathway .
Ciliopathies are a group of autosomal recessive disorders caused by a dysfunction of the primary cilium . These conditions are multisystemic disorders , affecting left-right symmetry ( situs inversus ) and various organs such as retina ( retinitis pigmentosa , Senior-Løken syndrome ) , brain ( cerebellar vermis aplasia , Joubert syndrome ) , liver ( cysts , intrahepatic biliary fibroadenomatosis ) , pancreas ( cysts ) as well as skeleton ( cone shape epiphysis , narrow thorax , polydactyly ) , and/or kidney ( renal cystic dysplasia ( RCD ) , nephronophthisis ( NPH ) ) [1 , 2] . RCD and NPH are major genetic causes of end stage renal failure in children and perinatal death for RCD . RCD is a kidney developmental defect whose antenatal diagnosis by ultrasound examination reveals hyperechogenic kidneys . Phenotypes range from enlarged cystic dysplastic kidneys to undersized , hypodysplastic kidneys . RCD is usually classified among the spectrum of CAKUT ( congenital anomalies of the kidney and urinary tract ) . NPH is characterized by atrophic kidney tubules with thickened basal membrane , interstitial fibrosis and , at a later stage , the development of cysts at the cortico-medullary junctions . Kidney size can be normal or reduced . The primary cilium is a microtubule-based antenna-like structure present at the cell surface of almost all vertebrate cells , which controls signaling pathways ( Hedgehog , canonical Wnt and planar cell polarity ( Wnt/PCP ) ) with a major role during development and homeostasis of the kidney and other organs . In renal tubular cells , the primary cilium functions as a mechano/chemo-sensor regulating cell cycle and PCP in response to urine flow , in order to control the orientation of the mitotic spindles along the axis of the elongating tubules and the organization of the epithelial cells with respect to their neighbors in the tissue . Defects in these processes result in cyst formation [3] . Ciliopathies are genetically heterogeneous diseases and mutations in >100 genes encoding ciliary proteins have been identified in affected patients [4] . Genotype-phenotype correlation analyses revealed that different mutations in the same gene could result in phenotypes with varying severity . Among these genes , missense mutations in NEK8/NPHP9 have been reported to lead to early onset isolated NPH [5] . However , a homozygous nonsense NEK8 mutation leading to absence of the protein was also identified in a family with three fetuses presenting with a more severe phenotype similar to Ivemark I and II syndromes , characterized by enlarged cystic dysplastic kidneys , pancreas and liver , associated with skeletal abnormalities , asplenia and congenital heart defects [6] . NEK8 is a serine/threonine kinase composed of an N-terminal kinase domain and five C-terminal Regulator of Chromosome Condensation 1 ( RCC1 ) repeat domains that belongs to the family of Never in Mitosis gene A ( NIMA ) proteins involved in the control of cell cycle progression [7] . In the cilium , NEK8 is located at the “Inversin ( INVS ) compartment” , a specific subcompartment of the proximal part of the axoneme , distal to the transition zone [8] . The function of this compartment is poorly understood , but human or mouse mutations in genes encoding components of the INVS compartment , INVS/NPHP2 , NPHP3 and ANKS6/NPHP16 , are known to lead to infantile nephronophthisis with cystic kidneys , congenital heart defects and laterality defects [9–13] . Additionally , NEK8 is also present in the nucleus where it controls the replication fork progression during S-phase and regulates DNA damage response [14] . Recently , NEK8 has been proposed as a regulator of the Hippo signaling pathway [15] . The Hippo pathway regulates organ size by controlling the balance between cell proliferation and cell cycle arrest through the phosphorylation state and nuclear shuttling of transcriptional co-factors YAP/TAZ [16 , 17] . Phosphorylation and nuclear shuttling of YAP/TAZ are strictly correlated to cell density , cell polarity and cellular actin cytoskeleton organization [18] . In low cell density conditions , YAP/TAZ are mainly unphosphorylated and able to translocate into the nucleus , resulting in cell proliferation . Conversely , at high cell density , YAP/TAZ are mainly phosphorylated and retained in the cytoplasm , leading to proliferation arrest . NEK8 has been reported to favor TAZ nuclear translocation , a process that is enhanced by NPHP4 , encoded by another gene causing NPH , highlighting these two proteins as inhibitors of the Hippo pathway [15 , 19] . Here , we report novel NEK8/NPHP9 mutations in five unrelated cases with severe multisystemic phenotypes . This study highlights the dual phenotype associated with the nature of the mutations and the key functions of NEK8 in ciliogenesis and cell proliferation/differentiation through regulation of YAP .
To identify novel mutations responsible for renal ciliopathies , we performed exon-enriched NGS targeting 1 , 222 genes associated with cilia structure/function , including all genes already known to be associated with ciliopathies ( “ciliome sequencing” ) [20–22] in two distinct cohorts of affected individuals: 342 patients with isolated or syndromic NPH and 200 fetuses or neonatal death cases with syndromic cystic dysplasia , including Meckel and Ivemark syndromes . Eight novel recessive mutations were identified in NEK8/NPHP9 in five unrelated families with severe overlapping phenotypes ( Fig 1A , Table 1 ) . All five cases presented with kidney involvement associated with extra-renal defects including situs inversus ( 4 cases ) , cardiomegaly ( 3 cases ) , paucity of bile ducts ( 3 cases ) , pancreas defects ( 3 cases ) , narrow thorax and short bowed femurs ( 2 cases ) , and brain defects such as corpus callosum or vermis agenesis ( 2 cases ) ( Table 1 ) . Patients/fetus from families 1 , 2 and 3 shared developmental abnormalities including asymmetric renal hypodysplasia with one absent or extremely reduced in size kidney and the other one with major dysplasia , diffuse interstitial fibrosis and incomplete cortico-medullary differentiation , cardiac septal hyperplasia and liver alterations including paucity of bile ducts ( Fig 1B and Table 1 ) . The fetus from family 3 , diagnosed with Ivemark I syndrome , presented with short pancreas and asplenia in addition to the defects listed above . The patient from family 2 and his affected sibling have already been reported [23] . In contrast , the fetuses from families 4 and 5 exhibited enlarged multicystic kidneys , cystic pancreas and liver , and agenesis of the vermis ( fetus 4 ) . The patient from family 1 carried compound heterozygous missense mutations: a paternally-inherited c . 259A>G variation in exon 3 leading to a missense alteration in the kinase domain ( p . Thr87Ala ) and a maternally-inherited c . 1804C>T variation in exon 13 resulting in an amino acid change in the RCC1 domain ( p . Arg602Trp ) ( Figs 1A , 1C and S1A , Table 1 ) . The patient from family 2 carried a homozygous missense mutation in exon 13 ( c . 1738 G>A ) leading to an amino acid substitution ( p . Gly580Ser ) in the RCC1 domain ( Figs 1A , 1C and S1A , Table 1 ) . The fetus from family 3 carried a maternally-inherited heterozygous variant of the last base pair of exon 4 ( c . 618G>A ) resulting in the in-frame skipping of exon 4 and loss of 44 amino acids in the kinase domain ( p . Val163-Ala206del ) , associated with a paternally-inherited heterozygous variation ( c . 1246G>A ) leading to a missense mutation ( p . Gly416Ser ) in the RCC1 domain ( Figs 1A , 1C and S1A , Table 1 ) . All four missense mutations were predicted as damaging by PolyPhen-2 and SIFT . The fetus from consanguineous family 4 carried a homozygous mutation affecting the first base of the 5' essential splice site in intron 1 ( c . 47+1 G>A ) . This mutation , which is predicted to totally abolish intron 1 splicing , is thus expected to lead to an absence of the protein . Finally , the fetus from family 5 carried two compound heterozygous nonsense mutations ( paternally-inherited p . Arg127* and maternally-inherited p . Arg462* ) that may result in truncated proteins and/or RNA decay as indicated by NEK8 mRNA quantification ( S1B Fig ) . We first examined the impact of patient mutations on the targeting of NEK8 to the cilium as well as on ciliogenesis , using patient fibroblasts and mIMCD3 kidney cells . We obtained primary fibroblasts from skin biopsies of family1 and 5 affected cases F1 II . 1 and F5 II . 1 ( subsequently referred as PT1 and PT5 ) harboring compound heterozygous mutations p . T87A/p . R602W or p . Arg127*/p . Arg462* , respectively . While NEK8 was located in the proximal part of the ciliary axoneme in control ciliated fibroblasts , it was absent from cilia in PT1 fibroblasts carrying missense mutations ( Fig 2A and 2A’ ) , and was instead detected onto a cytoplasmic juxtanuclear vesicular compartment that we identified as the Golgi apparatus ( S2A Fig ) . As an aside , we noted that endogeneous NEK8 or transfected NEK8-GFP proteins were also present at the Golgi in control fibroblasts ( S2A and S2B Fig ) . However , in these cells , NEK8 Golgi localization seemed to be transient as it was clearly observed at low confluence ( S2C Fig ) and decreased in ciliated cells , suggesting a dynamic localization of NEK8 during ciliogenesis . In PT5 fibroblasts carrying nonsense mutations , no specific NEK8 staining was detected , indicating that the truncating mutations result in the loss of protein expression . Subsequent evaluation of the impact of NEK8 mutations on ciliogenesis showed that the percentage of ciliated cells was significantly decreased in PT1 fibroblasts compared to control cells ( 50% vs 80%; Fig 2B ) . In contrast , ciliogenesis was not affected in PT5 fibroblasts ( Fig 2B ) , in agreement with previously reported data showing that renal and MEF cells from Nek8 knockout mice do not show ciliogenesis defects ( 10 , 25 ) . In addition , cilia length was more reduced in cells with missense mutations than in those with loss-of-function mutations ( Fig 2C ) . In order to characterize the specific effect of each NEK8 missense mutation , we generated murine inner medullary collecting duct ( mIMCD3 ) cells depleted for Nek8 expression using a lentiviral vector for shRNA expression . A 80% reduction of the mRNA level was obtained in the Nek8 knock-down cell line ( shNEK8 ) compared to the control cell line transduced with the empty vector ( pLKO ) leading to a non detectable expression of the protein by immunofluorescence and western blot analyses ( S3A , S3B and S3D Fig ) . The shNEK8 cells were then stably transfected with human NEK8-GFP constructs encoding either the wild-type protein ( WT ) or the missense variants identified in the patients . In addition , the p . H425Y NEK8 variant , located in the RCC1 domain and previously associated with isolated infantile NPH , was used as a control as it has been reported to alter NEK8 localization at the cilium [5 , 24] . The level of expression of human NEK8-GFP fusions was in the same range in all the cell lines ( S3C and S3D Fig ) . The WT NEK8-GFP protein localized to the ciliary axoneme at the “inversin ( INVS ) compartment” ( Fig 2D and 2D' ) . The proteins with mutations in the RCC1 domains , p . G580S and p . R602W , were no longer present at the cilium and accumulated into the cytosol , similar to the p . H425Y protein [5] . This is in agreement with the major role of the RCC1 domains in cilia localization [24] . In contrast , an axonemal staining , weaker than in control cells , was detected in half of the ciliated cells expressing the NEK8 protein mutated in the kinase domain ( p . T87A ) ( Fig 2D and 2D' ) . This result is coherent with a previous study showing that some NEK8 mutations in the kinase domain and affecting the kinase activity do not prevent its localization to the cilia in cell culture and in mice [10] . However , the decreased localization of NEK8 T87A-GFP protein points to a role of the kinase domain , possibly independently of the kinase activity , in ciliary NEK8 targeting process . Then , we evaluated ciliogenesis in the various mIMCD3 cell lines . As observed in PT5 fibroblasts , ciliogenesis was not significantly affected in shNEK8 cells ( Fig 2E ) . However , expression of all but WT and p . H425Y NEK8-GFP proteins led to a reduced percentage of ciliated cells , similar to what we observed in PT1 fibroblasts ( p . T87A/p . R602W ) , highlighting a “gain of function” effect of the missense mutations on this process ( Fig 2E ) . We next addressed the effect of NEK8 mutations on the ciliary localization of its partner ANKS6 [11] . While ANKS6 was located at the INVS compartment in 58% of control ciliated fibroblasts , its staining at the cilium was completely lost in PT1 and PT5 fibroblasts ( Fig 3A and 3A’ ) . As previously reported [10] , a dramatic reduction of ANKS6 positive cilia was also observed in shNEK8 mIMCD3 cells ( 40% compared to 90% in control cells ) ( Fig 3B and 3B' ) . This phenotype was rescued by the re-expression of WT NEK8-GFP ( 80% of ANKS6-positive cilia ) , whereas none of the mutant forms were able to fully restore ANKS6 localization at the cilium ( Fig 3B' ) . Moreover , in ANKS6 positive cilia of NEK8 mutant expressing cells , ANKS6 localization was restricted to the base of the cilium and did not extend along the INVS compartment as in control or NEK8-WT expressing cells ( Fig 3B” ) . We next investigated the impact of NEK8 mutations on the interaction with ANKS6 . Co-immunoprecipitation experiments demonstrated that only the p . T87A mutation led to a significant decrease of ANKS6 binding ( Fig 3C and 3C’ ) . This observation is in accordance with the recent report showing that NEK8 interacts directly with ANKS6 via its kinase domain [10 , 11] . It is also in agreement with the observed lack of ciliary ANKS6 in shNEK8 cells expressing the p . T87A mutation . Altogether , these data demonstrate that all the NEK8 missense mutations affect both ciliogenesis and biogenesis of the INVS compartment in renal tubular cells and fibroblasts , by preventing correct targeting of NEK8 to cilia and/or binding to or recruitment of ANKS6 . As NEK proteins are involved in cell cycle regulation [7] , we sought to examine how NEK8 mutations affect the cell cycle and proliferation . Counting the cell number over 7 days of culture revealed that patient fibroblasts behave differently depending on the type of mutation . While the population of PT1 cells with the missense mutations failed to expand as fast as control cells , PT5 fibroblasts with loss-of-function mutations expand much faster ( Fig 4A ) . In order to characterize this difference , we analysed cell cycle progression of control and patient fibroblasts . At low cell confluence , Ki-67 staining , a marker of G1 to M phase , did not show any difference ( Fig 4B ) . However , analysis of cell cycle by flow cytometry revealed a higher proportion of cells in S-phase in fibroblasts from patients compared to control , at the expense of cells in G0/G1 ( S4 Fig ) . We also analysed Ki-67 staining at high cell confluence , when cells are ciliated . As expected , in this condition , most if not all control cells were Ki-67-negative , likely in quiescent G0 status . In contrast , ~20% of PT1 and PT5 cells remained Ki-67-positive , including ciliated cells ( Fig 4B and 4B’; insets ) , indicating that patient cells fail to enter into G0 . This was confirmed by flow cytometry analysis , showing an increased proportion of cells in G2/M for both patient cell lines ( S4 Fig ) . Altogether , these results indicate a dysregulation of the cell cycle in patient cells , likely resulting in increased cell proliferation . This is in agreement with the data obtained by counting PT5 cell number that increases over time; however , for PT1 fibroblasts , we observed by Annexin-V staining that they underwent more apoptosis than control and PT5 cells after 48 hours of culture ( 33% vs 15% of control and PT5 cells; Fig 4C and 4C' ) , explaining why the population PT1 do not expand as well as PT5 fibroblast cells . NEK8 , like other proteins involved in ciliopathies , has recently been described as a regulator of DNA damage response ( DDR ) , with loss of NEK8 dramatically affecting S-phase progression upon DNA stress conditions [14] . Unlike the mentioned report , in non stress conditions we detected a significant increase in the proportion of nuclei positive for γH2AX ( phosphorylated form of H2AX ) , a marker of DNA double-strand breaks [25] , in PT1 fibroblasts compared to control and PT5 cells ( 40% vs 20% positive cells , respectively ) ( Fig 4D and 4D’ ) . The increase of γH2AX staining was also detected in kidney sections from the same patient , compared to control kidney ( Fig 4E ) . Hence , NEK8 missense mutations cause defects in DNA repair , which may lead to apoptosis during cell proliferation . In order to study the impact of NEK8 mutations on epithelial organization , we performed 3D matrigel culture on mutated NEK8 re-expressing mIMCD3 cells [26] . As previously reported [14] , 20% of the structures formed by shNEK8 cells were abnormal , without a clearly formed lumen ( multi-cellular aggregates , tubular structures with excessive branching , malformed spheroids ) after 5 days of culture ( Fig 5A and 5A’ ) . However , the most striking and previously unreported phenotype consisted of shNEK8 enlarged single lumen spheroids ( Fig 5A , 5A’ and 5A” ) . These enlarged structures correlated with a sustained Ki-67 staining ( Fig 5B and 5B’ ) during spheroid formation , as observed in patient fibroblasts in high cell density . Indeed , while at 2 days of culture the majority of control and shNEK8 spheroids were Ki-67-positive , at 5 days of culture only 10% of the structures remained Ki-67-positive in the control compared to 58% in shNEK8 spheroids . Expression of NEK8-WT protein , as well as p . H425Y form , partially rescued the proportion of normally-sized spheroids ( Fig 5A’ and 5A” ) . In contrast , expression of the three other missense mutated proteins led to an increased proportion of malformed or tubular-shaped structures ( p . G580S , p . R602W ) ( Fig 5A’ ) or failed to restore the correct size of the spheroids ( p . T87A ) ( Fig 5A” ) . These results indicate that in vitro the loss of function of NEK8 leads to an overgrowth of the spheroids ( enlarged spheroid structures ) , whereas the RCC1 missense mutations identified in the severe cases ( p . G580S and p . R602W ) affect epithelial morphogenesis , which may reflect the pathogenic effect of the different types of NEK8 mutations during kidney development . Since NEK8 has been reported as a regulator of the Hippo pathway , a critical pathway controlling growth and organ size [15] , we investigated the impact of NEK8 mutations on the regulation of this pathway . For this , we analyzed YAP localization and phosphorylation in low and high cell density , reflecting proliferating ( YAP active ) and quiescent ( YAP inactive ) states . As expected [18 , 27] , in control fibroblasts , YAP strongly accumulated into the nucleus in the low cell density condition ( non ciliated cells ) , whereas it was hardly detectable in the nucleus when cells reached high cell density and became ciliated ( Fig 6A , insets and Fig 6A’ ) . In contrast , both patient cell lines present altered nuclear YAP staining . PT1 fibroblasts showed a weak nuclear YAP staining compared to control cells at low cell density and this expression level was not modulated by confluence ( Fig 6A and 6A’ ) . In contrast , a strong nuclear YAP staining was observed in PT5 fibroblasts at low cell density , that decreased at high cell density but remained higher than in control cells ( Fig 6A and 6A' ) . Consequently , whereas the vast majority of control ciliated cells did not show any nuclear YAP staining ( Fig 6A ) , maintenance of nuclear YAP staining occurred in about 40% of ciliated patient fibroblasts ( Fig 6A , insets and Fig 6A” ) . These results indicate that NEK8 mutations lead to an improper regulation of nuclear YAP shuttling from proliferation to quiescent state , which may affect cell growth and differentiation . Upon cell confluence , inactivation of YAP is mediated by its phosphorylation on Ser127 [17] . As expected , immunofluorescence analysis revealed that cytosolic phospho-YAP increased with cell density in control fibroblasts ( S5 Fig ) . A similar increase was observed in patient fibroblasts , indicating that NEK8 mutations do not affect YAP phosphorylation ( S5 Fig ) . YAP phosphorylation is regulated by the MTS1/2-SAV1 complex , which has also been recently reported to promote ciliogenesis and to localize at the basal body in ciliated cells [28] . We thus carefully examined phospho-YAP localization and detected the presence of phospho-YAP along the axoneme in 80% of cilia in control and PT5 fibroblasts ( Fig 6B and 6B’ ) . However , in PT1 fibroblasts , phospho-YAP was absent from cilia and the staining was restricted to the cilium base in half of the ciliated cells ( Fig 6B and 6B’ ) . Altogether , these data underline that NEK8 mutations differently impair the cell density regulated nucleocytoplasmic shuttling of YAP , whereas only the missense mutations alter the localization of phospho-YAP at the cilium . In order to better understand how NEK8 mutations affect YAP nucleocytoplasmic shuttling , we analysed the ability of WT and mutated NEK8-GFP to promote nuclear YAP-myc localisation in co-transfected HEK393 cells . The co-transfection of each NEK8 mutant form with YAP-myc decreased the nuclear translocation of YAP-myc compared to WT NEK8-GFP , confirming the results observed in PT1 fibroblasts ( S6A Fig ) . Moreover , proximity ligation assay performed on cells co-expressing YAP-myc and NEK8-GFP-WT revealed that the two proteins are in close vicinity and are likely to interact at the perinuclear region ( S6B Fig ) , supporting a direct role of NEK8 in YAP regulation . Then , in order to examine if nuclear YAP imbalance had an impact on target gene regulation , we analysed the expression of YAP target genes as well as the transcriptional YAP co-regulator TEAD4 in control and patient fibroblasts . In control cells , the expression of CYR61 , CTGF and TEAD4 was decreased in confluent cells versus non-confluent cells ( Fig 6C ) , thus following the amount of nuclear YAP in these cells , as previously described [29] . In contrast , in PT1cells the expression of these genes was maintained at a similar level in high versus low confluent cells , consistent with maintenance of nuclear YAP localization in confluent cells . In PT5 cells , the expression of YAP target genes also reflected the nuclear YAP localization , with a high level of expression in non-confluent cells that decreased when cells reached confluence , although remaining at a higher level than in control cells . Among the signalling pathways reported to be downstream of YAP , we examined the Notch pathway , crucial for kidney and liver development [30 , 31] . In agreement with the upregulation of this pathway upon cell-cell contact , we observed an overexpression of JAG1 as well as the downstream target HES1 at high cell confluence in both control and PT1 fibroblasts . However , this increase was much higher in patient cells ( Fig 6D ) , indicating that dysregulation of nuclear YAP may also affect Notch signaling . In order to investigate if NEK8 mutations also led to YAP activation in vivo , we studied Yap and target gene expression in mutant juvenile cystic kidney ( Jck ) mice which bear a missense p . G448V mutation in the highly conserved RCC1 domain of Nek8 [32 , 33] . Immunohistochemistry assay showed that Yap expression was predominantly cytoplasmic in the kidneys of wild-type mice , whereas it was markedly increased in nuclei of kidneys of 5-week old Jck mice , an age at which these mice exhibited a polycystic kidney disease ( Fig 7A ) . Yap staining was particularly intense in the nuclei of tubular epithelial cells lining the cysts ( Fig 7A , insets ) . Western blot analysis of whole-kidney extracts ( Fig 7B and 7B' ) confirmed that the expression of both Yap and phospho-Yap ( S127 ) was increased in mutant mice . Nevertheless , the phospho-Yap ( S127 ) / Yap ratio was decreased , pointing again to an upregulation of Yap activity . In line with these observations , quantitative RT-PCR confirmed that Yap target genes , i . e . Ctgf , Cyr61 , Ankrd1 and Birc5 were upregulated in Jck mice ( Fig 7C ) . To determine if the epithelial morphogenesis abnormalities observed in shNEK8 and mutated NEK8 re-expressing cells were caused by deregulation of the Hippo pathway , we quantified nuclear YAP staining during spheroid formation . Control and shNEK8 cells grown in 3D in matrigel were fixed after 2 , 3 and 5 days of culture and stained for YAP ( Fig 8A ) . After 2 days , the majority of control and shNEK8 spheroids were positive for YAP . In control cells , the proportion of YAP-positive spheroids dramatically decreased to 15% at day 5 ( Fig 8A and 8A’ ) . However , nuclear staining of YAP was still present in 75% of shNEK8 spheroids after 5 days of culture . The continuous activation of YAP in the nucleus in shNEK8 mIMCD3 cells could thus promote cell growth in forming structures , causing abnormal enlarged spheroids as recently described in MDCK cells [34] . To confirm this hypothesis , we performed size rescue experiments using Verteporfin , an inhibitor of YAP-TEAD4 interaction [35] . Indeed , we observed that 1 μM of Verteporfin caused a reduction in size of the spheroids formed by shNEK8 cells after 5 days ( Fig 8B ) . In parallel , we confirmed that the increased expression of YAP targets observed in patient fibroblasts and shNEK8 IMCD3 cells was reduced with Verteporfin treatment ( S7 Fig ) . We also investigated if persistence of nuclear YAP in confluent patient fibroblasts was involved in the abnormal activation of the Notch pathway . As shown in Fig 8C , Verteporfin treatment performed in the high cell density condition dramatically reduced JAG1 expression in patient fibroblasts , further demonstrating the link between YAP activation and Notch dysregulation in NEK8 mutant cells . Finally , we examined the impact of NEK8 mutations identified in patients in zebrafish , an in vivo model relevant for ciliopathies [36] . Embryos injected with nek8 morpholino ( MO ) displayed the classical ciliopathy-related phenotype including curved body axis , laterality defects and pronephric cysts ( Fig 9A and 9A’ and Fig 8A and 8A’ ) , as previously described [37] . Body curvature was partially rescued by co-injection of human WT NEK8-GFP RNA ( 43% of normal embryos compared to 20% in nek8 morphants; Fig 9A’ ) but not by mutated NEK8-GFP RNA ( p . T87A and p . R602W ) , thus confirming the pathogenicity of the human missense mutations . Of note , we observed that co-injection of nek8 MO with WT NEK8-GFP RNA led to shortened dorsally curved embryos with occasionally a unique centered eye ( Fig 9A and 9A’ ) , a phenotype that was exacerbated by NEK8 missense mutations ( 60% vs 30% ) , further indicating their gain of function effect . Overexpression of human NEK8 accounts for the shortened dorsally curved phenotype since it was observed in 40% of embryos injected with WT RNA only ( Fig 9A” ) . We also observed laterality defects ( 70% of embryos ) and pronephros abnormalities ( cysts or developmental defects in 50% of embryos ) upon human NEK8 overexpression ( S8A and S8A’ Fig ) . As a similar dorsal curvature phenotype has been reported for embryos injected with human YAP RNA [38] , we performed rescue experiments using Verteporfin treatment ( Fig 9A” , S8 Fig ) . WT NEK8-GFP RNA-injected embryos were treated with 20 μM Verteporfin from 90% epiboly stage to 34 hours post fertilisation ( hpf ) . Analysis of Yap target gene expression by qPCR revealed that human NEK8 overexpression does induce an upregulation of the target genes , which is blocked by Verteporfin treatment ( S8B Fig ) . Contrary to laterality defects which remained unchanged , the proportion of stunted dorsally curved embryos and pronephros abnormalities decreased by 50% and 25% respectively upon treatment ( Fig 9A” , S8A” Fig ) . These data indicate that the NEK8 overexpression-related phenotype partially results from an upregulation of YAP activity in zebrafish . Altogether , these data demonstrate that abnormal YAP activation accounts for the epithelialisation , signaling and morphogenesis defects linked to NEK8 mutations .
To date , only two recessive human NEK8 mutations had been reported , one missense mutation in the RCC1 domain in a patient with early onset NPH and one nonsense mutation in the same domain in three fetuses from a consanguineous family with Ivemark I/II syndromes including cystic dysplastic lesions occurring in kidneys , liver and pancreas , and heart and skeletal defects [5 , 6] . Here , we describe 8 novel NEK8 mutations in five cases with severe multi-organ developmental defects , and the first association of NEK8 mutations with renal hypodysplasia and agenesis , situs inversus , agenesis of the vermis and bile duct paucity . Based on our results , NEK8 seems to be a major gene for renal dysplasia , since mutations were identified in 5 out of 200 analyzed families with dysplastic kidneys . Conversely , the previously identified mutation in a patient with infantile NPH seems to be a rare event , as we did not identify any other NEK8 mutation among the 342 analyzed NPH families . We also report the first two human mutations in the serine/threonine kinase domain of the protein . We observed a strong genotype-phenotype correlation . Fetuses with total NEK8 loss-of-function mutations ( c . 47+1G>A , family 3; p . R127*/p . R462* , family 5 ) presented enlarged cystic kidneys and pancreas associated with proliferative cystic biliary ducts , characteristics of Renal-Hepatic-Pancreatic Dysplasia syndrome ( OMIM #208540 ) , as described for the previously reported fetuses with a nonsense mutation [6] . In contrast , the three patients carrying missense mutations ( p . T87A , p . R602W , p . G580S and p . G416S ) or in-frame deletion due to a splicing defect ( p . V163-A206del ) presented with asymmetric dysplasic/hypodysplasic kidneys ( agenesis in one case ) with loss of differentiation , cortical interstitial fibrosis , dilated tubules and cartilage nodules , associated with paucity of bile ducts ( Table 1 ) . The functional analyses of the NEK8 mutations indicate that loss-of-function and missense mutations differentially alter ciliogenesis , proliferation/apoptosis and epithelial morphogenesis . Indeed missense mutations exacerbate some of the defects due to NEK8 loss of function both in vitro and in vivo ( zebrafish ) , highlighting their likely gain of function effect . In particular , only missense mutations lead to prominent ciliogenesis defects with reduction of percentage of ciliated cells and cilia length in fibroblasts and mIMCD3 cells . While both types of mutations affect cell cycle regulation , missense mutations also alter the function of NEK8 as a regulator of DNA damage response [14] , resulting in increased cell apoptosis in fibroblasts and kidney tissue . 3D culture assays showed that shNEK8 mIMCD3 cells ( loss of function ) form large spheroids ( cystic phenotype ) compared to NEK8 mutant form re-expressing mIMCD3 cells that mostly fail to get organized into spheroids ( dysplastic phenotype ) . Finally , co-injection of human mutant NEK8 RNA in nek8 zebrafish morphants further enhances their severe morphological alterations , resulting in a shortened dorsally curved body axis . It is noteworthy that the missense mutation previously reported in a patient with infantile NPH ( p . H425Y , [5] ) does not have the same gain of function effect , thus explaining the less severe phenotype . Therefore , this genotype/phenotype correlation points out the dual function of NEK8 for which a loss of function ( nonsense mutations ) leads to proliferative/cystic phenotypes and a gain of function ( missense mutations ) to hypodysplastic phenotypes with loss of differentiation ( S9 Fig ) . Although Nek8 mouse model phenotypes are different from those of human cases , a genotype-phenotype correlation seems to also exist in rodents in term of severity of the renal phenotypes . The Nek8jck/jck mice , carrying a missense homozygous mutation in the RCC1 domain ( p . G448V ) , which has been demonstrated to be a gain of function , develop enlarged cystic kidneys [32] . In contrast , Nek8 knockout mice ( Nek8tm1Bei ) present a mild renal phenotype , with dilated proximal tubules and glomerular cysts [39] and the mouse model with a missense mutation ( p . I124Y ) in the kinase domain ( Nek8roc ) exhibits hydroureter , cystic tubular dilations and small glomerular cysts [10] . However , both Nek8tm1Bei and Nek8roc mice also present situs inversus and heart defects that lead to death at birth and which may prevent renal cyst formation during the final , post-birth steps of murine nephrogenesis . Association of kidney defects with situs inversus and heart defects was also observed in the five human cases . This phenotype is consistent with a general alteration of the ciliary function of NEK8 and the integrity of the INVS compartment ( absence of NEK8 or ANKS6 proteins in the cilia ) . Indeed , mutations in genes encoding other components of the INVS compartment ( INVS/NPHP2 , NPHP3 and ANKS6/NPHP16 ) are known to lead to infantile NPH associated with enlarged cystic kidneys or to kidney cystic dysplasia associated with congenital heart defects and situs inversus [9 , 11–13] . Interestingly , we also identified a homozygous frameshift mutation ( c . 1010_1011del , p . G337Afs*16 , family 6 ) in ANKS6 ( S10 Fig ) in a fetus whose phenotype was similar to that of NEK8 loss of function cases , i . e . enlarged cystic kidneys associated with enlarged fibrotic pancreas , situs inversus and cardiopathy . This is in agreement with the phenotype of Anks6Streaker mice whose Anks6 mutation ( p . M187K ) decreases the binding to and activation of Nek8 and leads to cystic kidneys , situs inversus and congenital heart defects , thus mimicking the phenotype of patients with NEK8 mutations [10] . Identification of this ANKS6 mutation , together with our functional data , strengthens the close relationship between NEK8 and ANKS6 , i . e . NEK8 recruits its target ANKS6 to the cilium , which in return enhances NEK8 kinase activity [10] . Besides its function at cilia , NEK8 is critical for cell cycle regulation . We demonstrate that NEK8 mutations lead to defective Hippo pathway regulation , with a decreased amount of nuclear YAP in proliferating cells ( missense mutations ) as well as maintenance of a pool of nuclear YAP in confluent ciliated cells ( missense and loss-of-function mutations ) in vitro in patient fibroblasts . Such defects were also observed in shNEK8 mIMCD3 spheroids and in vivo in Jck cystic tubular cells . Maintenance of nuclear YAP in patient confluent ciliated cells was accompanied by higher expression of YAP targets compared to controls , sustained Ki-67 staining and abnormal cell cycle with an increased proportion of cells in G2/M . Therefore , NEK8 mutant cells still undergo proliferation and fail to differentiate when reaching confluence . Moreover , we show that cells harboring NEK8 missense mutations are more subject to DNA damage than NEK8 defective cells and undergo apoptosis , which can contribute to the difference in cell growth resulting in hypodysplastic versus enlarged multicystic kidneys respectively ( S9 Fig ) . Several mechanisms could account for nuclear YAP misregulation . Association of NEK8 with TAZ , the other major Hippo pathway effector , favors TAZ translocation into the nucleus [6 , 15] . We show that NEK8 interacts with YAP in a perinuclear region , suggesting a similar regulation process for YAP and TAZ . Moreover , we show that NEK8 missense mutations alter YAP nuclear translocation . As all mutations in both kinase and RCC1 domains affect NEK8 localization into the nucleus in mIMCD3 cells ( S3E and S3E’ Fig ) , NEK8 mutant forms might thus affect YAP shuttling into the nucleus , through their own defective nuclear translocation . NPHP4 has previously been reported to favor NEK8 and TAZ translocation into the nucleus [15] . We can hypothesize that NEK8 mutations affect binding to NPHP4 , resulting in a less efficient translocation of NEK8 , and consequently YAP into the nucleus . During normal cell differentiation , ciliogenesis and quiescence are accompanied by activation of the Hippo pathway ( i . e . YAP inactivation ) and proteasome-mediated degradation of cytoplasmic NEK8 [24] . However , the constant level of NEK8 protein expression detected in low and high confluent cells of PT1 with missense mutations suggests that these mutations preserve NEK8 from degradation , resulting in maintenance of nuclear YAP and consequently the lack of proper cell differentiation . This result is in agreement with our in vivo observations showing that NEK8 overexpression in zebrafish embryos mimics the YAP overexpression phenotype [38] . Moreover , the preserved YAP nuclear localization in cells with NEK8 loss-of-function mutations indicates that other proteins help YAP to translocate into the nucleus in proliferative condition , but also control its partial downregulation in confluent cells . This may explain why loss-of-function NEK8 mutations partially preserve nephrogenic differentiation , evidenced by the presence of some mature glomeruli in fetal renal biopsies . The maintenance of YAP in the nucleus in confluent patient fibroblasts and renal cells in 3D culture may also be associated with defective Hippo pathway activation at the cilium in NEK8 mutant conditions . Recent studies on MST1/2 , two major activators of the Hippo pathway , showed that they localize at the basal body and promote ciliogenesis [28] . In this study , we report the presence of phospho-YAP at the cilium and that this localization is partially affected in the presence of NEK8 missense mutations ( S9 Fig ) . Phospho-YAP at the cilium may thus be a key component of activation of Hippo pathway under the control of NEK8 . Finally , soft matrix , such as matrigel in 3D culture assays , is known to promote cytoplasmic retention of YAP and TAZ resulting in limited cell growth , via LATS independent non-canonical Hippo pathway activation , involving cytoskeleton , cell junctions , RhoGTPases or GPCR signaling [40] . Maintenance of nuclear YAP in enlarged spheroids in the absence of NEK8 suggests that NEK8 could also regulate YAP through non-canonical mechanisms . The Hippo pathway is a highly integrative pathway whose regulation is connected to that of many other signalings crucial for organogenesis , including Wnt/β-catenin , TGF-β , BMP and Notch . YAP dysregulation due to NEK8 mutations is thus expected to play a major role in the development of the multisystemic defects presented by the patients/fetuses . Specific inactivation of Yap in the nephrogenic lineage ( YapCM-/- ) leads to a reduced number of nephrons [41] , as seen in the patients with NEK8 missense mutations . We can hypothesize that an abnormal amount of nuclear YAP during kidney development , as shown in patient fibroblasts and Nek8jck/jck mice , would affect the expression of downstream targets that might participate to the pathophysiological processes . CYR61 , a gene expressed in tubules and glomeruli in fetal kidney , encodes a CCN protein that interacts with integrins to mediate cell adhesion , migration and differentiation during nephrogenesis . Upregulation of the Notch1 ligand JAG1 may induce a dysregulation of Notch pathway , required during nephron tubular development [42 , 43] , thus contributing to renal hypodysplasia . Finally , CTGF , encoding a regulator of cartilage morphogenesis and mediator of fibrosis [44 , 45] could be involved in the formation of cartilage islets and fibrosis observed on most of the kidney biopsies of NEK8 mutated patients/fetuses . In the liver , YAP is highly expressed in bile ducts and regulates the Notch pathway for ductal specification during development [31] . In the pancreas , YAP-TEAD regulates the transcriptional network controlling pancreatic cell proliferation and differentiation [46] . As for kidney defects , liver and pancreas defects vary according of the type of NEK8 mutation , suggesting that NEK8 loss of function and gain of function differentially affect the hepatic/pancreatic transcriptional program , leading to either proliferation or paucity of bile ducts in the liver , and cysts or fibrosis in the pancreas , respectively . Finally , Hippo pathway dysregulation is associated with heart overgrowth in mice [47] , thus highlighting the involvement of this pathway in the pathophysiological mechanisms leading to cardiac defects seen in the NEK8 patients/fetuses . In conclusion , we demonstrate that NEK8 is a multifunctional protein whose alterations lead to severe developmental abnormalities due to the synergic effect of dysfunction of key processes and signaling pathways . The demonstration of the central role of YAP dysregulation in NEK8 mutant conditions highlights potential therapeutic targets for the patients .
This study was conducted with the approval of the « Comité de Protection des Personnes pour la Recherche Biomédicale Ile de France II » . Approval was obtained under numbers 2007-02-09/DC-2008-229 and 2009-164/DC-2011-1449 ( fetuses ) and 2008-A01039-46/DC-2008-229 ( nephronophthisis patients ) . For each patient/fetus , written informed consent was obtained from the parents . For studies using animal data: housing and handling of mice were performed in accordance with the guidelines established by the French Council on animal care "Guide for the Care and Use of Laboratory Animals": EEC86/609 Council Directive—Decree 2001–131 . The project was approved by the departmental director of "Services Vétérinaires de la Préfecture de Police de Paris" and by the ethical committee of the Paris Descartes University ( approval number: A75-15-34 ) . 342 patients with isolated or syndromic NPH and 200 fetuses or early neonatal death cases with syndromic cystic dysplasia , including Meckel and Ivemark syndromes , were studied . Genomic DNA was isolated from peripheral blood or frozen tissues using standard procedures . Ciliary exome targeted sequencing and bioinformatic filtering was conducted in affected individuals using a custom SureSelect capture kit ( Agilent Technologies ) targeting 4 . 5 Mb of 20 , 168 exons ( 1221 ciliary candidate genes ) , including NEK/NPHP9 . Briefly , Agilent SureSelect capture libraries were prepared from 3 μg of genomic DNA samples sheared with a Covaris S2 Ultrasonicator according to manufacturer’s instructions . The SOLiD molecular barcodes for traceable ID of samples were added at the end of the capture step . The Ovation Ultralow System ( NuGEN Technologies ) was used to prepare HiSeq2500 pre-capture barcoded libraries . The ciliome capture by hybridization was performed on a pool of 10 to 16 barcoded precapture libraries . Sequencing performed on SOLiD5500XL ( Life Technologies ) and HiSeq2500 ( Illumina ) was done on pools of barcoded ciliome librairies ( 64 barcoded ciliome libraries per SOLiD FlowChip and 16 ciliome libraries per lane of HiSeq FlowCell ) . Paired-end reads were generated ( 75 + 35 base reads for SOLiD , 100 + 100 base reads for HiSeq ) and mapped on human genome reference ( NCBI build37/hg19 version ) using Burrows-Wheeler Aligner ( Illumina ) or mapread ( SoliD ) . Downstream processing was carried out with the Genome Analysis Toolkit ( GATK ) , SAMtools , and Picard Tools , following documented best practices ( http://www . broadinstitute . org/gatk/guide/topic ? name=best-practices ) . All variants were annotated using a software system developed by the Paris Descartes University Bioinformatics platform . The mean depth of coverage obtained was greater than 90x , and more than 89% of the exome was covered at least 15x . Different filters were applied to exclude all variants located in non-exonic regions , pseudogenes , UTRs or known polymorphic variants with a frequency above 1% , i . e . present in databases such as dbSNP , 1000 genome projects and all variants identified by in-house exome sequencing ( 5150 exomes and 1020 ciliomes ) . The functional consequence of missense variants was predicted using SIFT ( http://sift . jcvi . org/www/SIFT_enst_submit . html ) and PolyPhen2 ( http://genetics . bwh . harvard . edu/pph2/ ) softwares . Control and affected individual fibroblasts were cultured in Opti-MEM supplemented with 10% fetal bovine serum , penicillin , streptomycin , uridine , sodium pyruvate and Ultroser G G serum substitute ( Pall Corporation ) . Control and patient fibroblasts ( 1 . 5 × 104 or 2 . 5× 104 cells respectively ) were plated on coverslips and grown for 2 days ( low confluence ) or 6 days followed by 48-hour serum deprivation ( high confluence ) before fixation . Murine inner medullary collecting duct ( mIMCD3 ) cells were cultured in DMEM F-12 and HEK293T in DMEM both supplemented with 10% fetal bovine serum , penicillin , streptomycin and L-Glutamine ( all from Life Technologies ) . For immunofluorescence , 2 . 5 x 104 cells were plated on coverslips and grown for 5 days before fixation . In all the experiments , the level of confluence was visually checked and counted to ensure similarity between control and samples . mIMCD3 and fibroblasts were fixed in 4% PFA in PBS 1X for 15 min followed by treatment with 50 mM NH4Cl for 15 min . Antibodies used for immunofluorescence were: NEK8 ( kind gift of D . Beier [32] ) , ANKS6 ( 1:50 , Sigma-Aldrich HPA008355 ) , GFP rabbit ( 1:500 , Life Technology A11122 ) , GFP chicken ( 2B Scientific , 1020 ) , GM130 ( 1:50 , BD 558712 ) , YAP ( 1:50 , Cell Signaling #4911 ) , phospho-YAP ( Ser127 ) ( 1:50 , Cell Signaling #4911 ) , γH2AX ( 1:500 , Millipore 05–636 ) and anti-acetylated α-tubulin ( 1:10000 , Sigma-Aldrich ) . Cells were permeabilized with Triton 0 . 5% for 10 min at room temperature and treated with blocking solution constituted of PBS 1X , 0 . 1% Tween 20 , 3% ( for fibroblasts ) or 1% ( for mIMCD3 ) BSA before incubating with primary antibodies overnight . Then , cells were washed 3 times with PBS 1X for 10 min and stained with appropriate Alexa Fluor-conjugated secondary antibodies ( 1:200 , Molecular Probes ) . Nuclei were stained with Hoechst . For Annexin-V assays , cells were first incubated with a cold solution constituted by 10 mM HEPES , 140 mM NaCl and 25 mM CaCl2 . Annexin V ( Life Technologies ) was secondly incubated in the same solution for 30 minutes at room temperature . Fixation was performed with PFA ( 4% ) for 20 minutes and nuclei were stained with Hoechst . Tissue biopsies embedded in paraffin blocks were sectioned ( 8 μm section thickness ) using a Leica microtome . Next , sections were immersed in xylene baths ( 5 minutes in the first bath , 5 minutes in the second bath ) , then rehydrated for 5 minutes in ethanol baths of decreasing concentrations ( 100% , 95% , 70% , and 40% ) and finally immersed in MilliQ water for 5 minutes . Dako target retrieval solution ( Dako ref . S1699 ) was used according to the manufacturer's instructions . The slides were blocked for 45 minutes at 4°C by 10% NDS ( Normal Donkey Serum ) diluted in PBT ( DPBS with 0 . 1% Triton X100 ) . Fluorescein label Peanut Agglutin ( PNA ) ( 1:200 , Vector Fl-1071 ) was used to detect collecting tubules . Other primary and secondary antibodies were used as described above . Slides were mounted in adapted medium , and analysed under an inverted confocal microscope Zeiss LSM 700 . Nek8-knockdown ( KD ) was performed in mIMCD3 cells by lentiviral infection of a shRNA expressing construct in pLKO puromycin vector ( Sigma-Aldrich sh1570 ) , as previously described [39] . Puromycin-resistant Nek8-KD cells were then transfected using Lipofectamine2000 with wild-type and mutated human NEK8-GFP constructs [24] and stable cell lines re-expressing NEK8-GFP were selected with double selection with geneticin and puromycin . NEK8-GFP variants were obtained through site-directed mutagenesis using Pfu turbo kit ( Invitrogen ) . YAP-myc construct has been described in [18] . HEK293 cells were transiently transfected using the calcium phosphate method . After 48 hours , cells were harvested with ice-cold PBS 1X . A small aliquot of this cell suspension was immediately removed and lysed directly in SDS-PAGE sample buffer as a whole cell lysate . The remaining harvested cells were lysed and treated in accordance with the Miltenyi-Biotec beads protocol . Protein dosage was performed using the BCA protein assay kit ( Thermo Scientific ) . Fifty micrograms of proteins were loaded on a 8% acrylamide gel ( Bio-rad ) , and Western blot was conducted using the indicated anti-FLAG M2 ( 1:1000 , Sigma-Aldrich F1804 ) , anti-GFP ( 1:1000 , Roche #1814460001 ) , anti-tubulin ( 1:10000 , Sigma-Aldrich T5168 ) . For in situ Proximity Ligation Assay ( PLA ) ( OLINK Biosciences , Uppsala Sweden ) , HEK293 cells were fixed 48h after transfection in 4% PFA for 15 min , permeabilized 10 min with PBS-0 . 1% Triton before treated with blocking solution , labeled with anti-rabbit GFP and anti-mouse Myc ( Thermo Fischer Scientific , #MS139P1 ) antibodies and then incubated with a pair of nucleotide-labeled secondary antibodies ( rabbit PLA probe MINUS and mouse PLA probe PLUS ) in hybridization solution . Interactions between the PLA probes , possible when within a distance less than 40 nm , were revealed by adding a ligase and by amplification of a rolling-circle product using labeled oligonucleotides and a polymerase , according to the manufacturer's instructions . Signals indicative of interactions were detected by confocal microscopy as fluorescent dots in visible red . Total cellular mRNA was isolated using Qiagen Extraction Kit and then treated with DNase I . 1 . 5 μg of total RNA was reverse-transcribed using Superscript II ( Life Technologies ) . Relative expression levels of genes of interest were determined by real-time RT-PCR using the Absolute SYBR Green ROX Mix ( ABgene ) and specific primers as follows: human NEK8 forward 5’-GCCTCAAGAGGGCTTTCGA-3’ and reverse 5’-AAGGTGCCACTCATGATCTTCAG-3’; mouse Nek8 forward 5'-GCACCTTGGCCGAGTTCAT-3' and reverse 5'-GCCAGCAGGATCTGCACAA-3'; human CTGF forward 5'-CGAAGCTGACCTGGAAGAGAA-3' and reverse 5'- GTACTCCCAAAATCTCCAAGCCT-3'; human CYR61 forward 5'-GAGTGGGTCTGTGACGAGGAT-3' and reverse 5'-GGTTGTATAGGATGCGAGGCT -3'; human TEAD4 forward 5'-GGACACTACTCTTACCGCATCC-3' and reverse 5'- TCAAAGACATAGGCAATGCACA-3; human JAG1 forward 5'-GCCGAGGTCCTATACGTTGC-3' and reverse 5'-CCGAGTGAGAAGCCTTTTCAA-3'; human HES1 forward 5'-TCAACACGACACCGGATAAAC-3' and reverse 5'-GCCGCGAGCTATCTTTCTTCA-3' . Experiments were repeated at least three times and gene expression levels were normalized to GAPDH . For qPCR analyses in IMCD3 cells , we used mouse primers described below . Experiments were performed on 5-week-old female mutant juvenile cystic kidney ( Jck ) mice bearing a Nek8 mutation ( The Jackson Laboratory ) and compared to wild-type littermates . Animals were fed ad libitum and housed at constant ambient temperature in a 12/12-hour light/dark cycle . For mouse samples , 4 μm sections of paraffin-embedded kidneys were submitted to heat-mediated antigen retrieval and incubated with antibody to Yap ( Cell Signaling Technology , 4912 , 1:100 ) , followed by a donkey anti-rabbit biotinylated antibody ( GE Healthcare ) at 1:200 . Biotinylated antibodies were detected using HRP-labeled streptavidin ( Dako ) at 1:2000 and 3–3′-diamino-benzidine-tetrahydrochloride ( DAB ) revelation . Western blot analyses were performed as previously described [48] . Briefly , protein extracts from kidneys were resolved by SDS-PAGE before being transferred onto the appropriate membrane and incubated with antibodies to phospho-YAP ( Ser127 ) ( Cell Signaling Technology , 4911 , 1:1000 ) , and YAP ( Santa Cruz , sc-101199 , 1:1000 ) , Gapdh ( Millipore , 1:5000 ) followed by the appropriate Alexa-conjugated secondary antibody ( Life Technologies ) . Fluorescence was acquired using a ChemiDoc MP Imaging System ( Bio-Rad ) , and densitometry was performed using Image Lab software 5 . 0 . For real-time RT-PCR , mRNA were extracted from whole kidney samples and Ctgf , Cyr61 , Birc5 and Ankrd1 expression were analysed by real-time RT-PCR using CFX96 Touch Real-Time PCR Detection System ( Bio-Rad ) . Primers ( Eurogentec ) were as follows: Ctgf forward 5’-GCTGACCTGGAGGAAAACATTAA-3’ and reverse 5’-TGACAGGCTTGGCGATTTTAG-3’; Cyr61 forward 5’-CCTTCTCCACTTGACCAGAC-3’ and reverse 5’-ATATTCACAGGGTCTGCCTTCT-3’; Birc5 forward 5’-CCCGATGACAACCCGATAGAG-3’ and reverse 5’-TGACGGGTAGTCTTTGCAGTC-3’; Ankrd1 forward 5’-CTGTGAGGCTGAACCGCTAT-3’ and reverse 5’-CCAGTGCAACACCAGATCCA-3’ . Rpl13 was used as the normalization control . A total of 7 . 5 x 104 cells/well were plated in triplicate in 6 well plates and grown for 1–7 days . Cells were incubated with complete medium as previously described . The number of cells was counted at the indicated time-points in triplicate . For flow cytometry analysis , cells were plated at a density of 1 x 105 cells/ml . The cells were pulse-labeled with BrdU for 30 min , washed with PBS , and treated with trypsin . Cells were fixed with ethanol and stained with anti-BrdU-FITC antibody ( BD Biosciences ) and propidium iodide , according to the manufacturer’s instructions . Flow cytometry analysis was carried out with the BD LSRII flow cytometry system and BD FACSDiva software . 96 well plates were coated with a thin layer of collagen ( collagen I , Rat Tail , Corning #354236 ) that was allowed to polymerize at 37° C for 30 minutes . 4 x 104 cells per well in the appropriate medium ( with antibiotics ) was mixed with Matrigel ( BD ) and allowed to polymerize at 37°C for 30 minutes . Subsequently , the appropriate medium was added and changed every 2 days . Samples were fixed after 2 , 3 or 5 days of culture . After two washes with PBS 1X , PFA 4% was added for 30 minutes . Antibody stainings were done as previously described; only incubation time with blocking solution was prolonged to 1 hour at room temperature . Zebrafish were maintained at 28 . 5°C under standard protocols . Tg ( cmlc2:GFP ) and Tg ( wt1b:GFP ) transgenic lines were used to assess heart looping and pronephros morphology , respectively . Control and nek8 ( ATG ) morpholinos [37] were injected into one-cell stage embryos at 0 . 4 pmol per embryo . Human full length NEK8-GFP RNA was obtained by in vitro transcription with mMESSAGE mMACHINE kit ( Ambion ) and injected into one-cell stage embryos at 100 pg per embryo . For Verteporfin treatment , embryos were injected with 100 pg of RNA and GFP-positive animals were selected at shield stage . Embryos were then treated with either DMSO or 20 μM Verteporfin from 90% epiboly stage to 34 or 52 hours post fertilization ( hpf ) , time points at which body curvature and laterality/pronephros phenotypes were measured , respectively . Phenotypes were analysed using a Leica M165FC stereoscope . For real-time RT-PCR , mRNA was extracted from whole embryos at 34 hpf by TRIZOL and ctgfa , reported to be specific to Yap unlike ctgfb [49] , cyr61 and tead4 expression were analysed by real-time RT-PCR using Absolute SYBR Green ROX Mix ( ABgene ) and specific primers as follows: ctgfa forward 5’-TCCTCACAGAACCGCCACCTTGCCCAT-3’ and reverse 5’-TCACGCCATGTCGCCAACCATCTTCTTGT-3’; cyr61 forward 5’-CCGTGTCCACATGTACATGGG-3’ and reverse 5’-GGTGCATGAAAGAAGCTCGTC-3’; tead4 forward 5’-AAGGAGGACTGAAGGAGCTGTTCGAGAAGG-3’ and reverse 5’-GCCGAATGAGCAGACTTTAGTGGAGGAGGT-3’ . gapdh was used as the normalization control . For treatment of human fibroblasts , Verteporfin ( Sigma-Aldrich , SML0534 ) was added after 5 days of culture when cells had achieved confluence . Several drug concentrations were tested and 0 . 5–0 . 075 μM were chosen as optimal non toxical conditions . For treatment of shNEK8 re-expressing NEK8-GFP mIMCD3 cells , we used drug concentration ranging from 0 . 5 to 4 μM . Verteporfin treatment was also used on control pLKO1 and shNEK8 mIMCD3 cells for rescue experiments in the matrigel 3D assay . In this case , the drug ( 1 and 2 μM ) was added after 2 days of culture and maintained until fixation at 3 or 5 days . | Genes mutated in ciliopathies encode proteins with various localizations and functions at the primary cilium . Here we report novel NEK8 mutations in patients with renal cystic hypodysplasia and associated ciliopathy defects . NEK8 belongs to a protein complex defining the Inversin compartment of the cilium . It is also a negative regulator of the Hippo signaling pathway that controls organ growth . We report genotype-phenotype correlation in the patients . We functionally demonstrate that the two types of mutations ( missense versus nonsense ) differentially affect ciliogenesis , cell apoptosis and epithelialisation . We also show that all the mutations lead to dysregulation of the Hippo pathway through nuclear YAP imbalance but that the nature of this imbalance is different according to the type of mutation . We confirm alteration of the Hippo pathway associated with Nek8 mutation in vivo in Jck mice . Remarkably , we show that morphogenesis defects observed in Nek8 knockdown epithelial cells or zebrafish embryos are rescued by Verteporfin , a specific inhibitor of YAP transcriptional activity , demonstrating the causative role of YAP dysregulation in the occurrence of these defects . Altogether , this study links NEK8 mutations to dysregulation of the Hippo pathway and provide molecular clues to understand the variability of the multiorgan defects in the patients . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"medicine",
"and",
"health",
"sciences",
"nuclear",
"staining",
"fibroblasts",
"mutation",
"developmental",
"biology",
"connective",
"tissue",
"cells",
"nonsense",
"mutation",
"embryos",
"kidneys",
"cellular",
"structures",
"and",
"organelles",
"research",
"and",
"analysis",
"methods",
"missense",
"mutation",
"specimen",
"preparation",
"and",
"treatment",
"embryology",
"staining",
"animal",
"cells",
"connective",
"tissue",
"biological",
"tissue",
"cell",
"staining",
"cell",
"biology",
"anatomy",
"cilia",
"genetics",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"renal",
"system"
] | 2016 | Novel NEK8 Mutations Cause Severe Syndromic Renal Cystic Dysplasia through YAP Dysregulation |
We sought to define protective mechanisms of immunity to Staphylococcus aureus and Candida albicans bloodstream infections in mice immunized with the recombinant N-terminus of Als3p ( rAls3p-N ) vaccine plus aluminum hydroxide ( Al ( OH3 ) adjuvant , or adjuvant controls . Deficiency of IFN-γ but not IL-17A enhanced susceptibility of control mice to both infections . However , vaccine-induced protective immunity against both infections required CD4+ T-cell-derived IFN-γ and IL-17A , and functional phagocytic effectors . Vaccination primed Th1 , Th17 , and Th1/17 lymphocytes , which produced pro-inflammatory cytokines that enhanced phagocytic killing of both organisms . Vaccinated , infected mice had increased IFN-γ , IL-17 , and KC , increased neutrophil influx , and decreased organism burden in tissues . In summary , rAls3p-N vaccination induced a Th1/Th17 response , resulting in recruitment and activation of phagocytes at sites of infection , and more effective clearance of S . aureus and C . albicans from tissues . Thus , vaccine-mediated adaptive immunity can protect against both infections by targeting microbes for destruction by innate effectors .
Staphylococcus aureus and Candida spp . are the second and third leading causes of bloodstream infections in hospitalized patients [1] . These organisms jointly cause at least 150 , 000 clinical bloodstream infections resulting billions of dollars of health-care expenditures and ∼40 , 000 deaths per year in the US alone [1]–[4] . Identification of immune mechanisms of protective adaptive immunity against these organisms is critical to lay the groundwork for development of active vaccine strategies against both organisms . We previously reported that vaccination with the recombinant N terminus of the candidal Als3p adhesin ( rAls3p-N ) with aluminum hydroxide ( Al ( OH ) 3 ) adjuvant improved the survival of mice subsequently infected intravenously with lethal inocula of Candida albicans or methicillin resistant Staphylococcus aureus ( MRSA ) [5]–[7] . The vaccine retained efficacy against both infections in B cell deficient animals but not T cell deficient animals [6] , [7] . Furthermore , adoptive transfer of CD4+ T cells but not B220+ B cells or immune serum improved the survival of recipient mice infected with both organisms [6] , [7] . Although T cells are necessary for rAls3p-N vaccine efficacy , lymphocytes are not capable of directly killing C . albicans or S . aureus in culture [8] , [9] . Therefore , the downstream effectors of vaccination against both organisms have remained unclear . In contrast to lymphocytes , phagocytes kill C . albicans and S . aureus in vitro [8] , [10] , [11] and in vivo [12]–[16] , especially when primed with pro-inflammatory cytokines such as interferon ( IFN ) -γ , which is produced by CD4+ lymphocytes . Therefore , we hypothesized that the end effectors of rAls3p-N vaccine-mediated protection against bloodstream infection caused by S . aureus and C . albicans were phagocytes primed by pro-inflammatory cytokines produced by vaccine-responsive lymphocytes . We sought to elucidate fundamental requirements of protective host immunity to bloodstream infection caused by S . aureus and C . albicans .
We previously established that the rAls3p-N vaccine was not effective against C . albicans iv infection in IFN-γ-deficient mice [6] . We sought to determine if IFN-γ was similarly required for vaccine-mediated protection against S . aureus , and also to determine if CD4+ T cells were the required source of IFN-γ production to mediate vaccine efficacy against both organisms . IFN-γ-deficient mice or their wild-type , congenic controls were vaccinated with rAls3p-N plus Al ( OH ) 3 ( vaccinated ) or Al ( OH ) 3 alone ( control ) , and boosted at three weeks . Two weeks following the boost , CD4+ splenic and lymph node lymphocytes from vaccinated or control donor mice were purified and cross-adoptively transferred to recipient mice ( IFN-γ deficient donor cells were transferred to wild type recipient mice , and visa versa ) . As a negative control , vaccinated or control IFN-γ knockout mice were infected without undergoing adoptive transfer . Mice were infected via the tail-vein with C . albicans or MRSA the day following adoptive transfer . IFN-γ-deficient mice receiving immune CD4+ lymphocytes from vaccinated , wild type donor mice had improved survival after either infection , whereas wild type mice receiving immune CD4+ lymphocytes from IFN-γ-deficient , vaccinated donor mice did not have improved survival ( Fig . 1 ) . Cells from control donor mice were not effective at improving survival of recipient mice . Hence , IFN-γ produced by vaccine-primed CD4+ T cells was required for mediating adaptive immunity against both infections . Because lymphocyte derived pro-inflammatory cytokines , including IFN-γ , can activate phagocytes to mediate superior killing of C . albicans or S . aureus in culture [8] , [11] , [17]–[22] , we sought to define the role of downstream phagocytes in adaptive immune-mediated protection . First we vaccinated mice as above and administered cyclophosphamide to induce neutropenia two weeks after the boost . Two days later we infected the mice with one of two clinical isolates of C . albicans , or with MRSA . The second isolate of C . albicans ( 15563 ) was used because it results in a less rapidly lethal infection than SC5314 , and the diminished severity of infection would afford the opportunity to unmask any subtle , residual protection afforded by vaccination in the neutropenic mice . Cyclophosphamide-induced neutropenia disrupted the improvement in survival mediated by the vaccine against infections caused by either strains of C . albicans and S . aureus ( Fig . 2 and Fig . S1 for the 15563 strain ) . We also tested vaccine efficacy in gp91phox−/− deficient mice , the phagocytes of which are unable to generate superoxide and have marked defects in microbial killing . Such mice have been previously shown to have enhanced susceptibility to pulmonary and intraperitoneal infection by C . albicans [23] , [24] , but have not been studied in the intravenous model . We performed pilot studies to determine how susceptible to iv infection with C . albicans or S . aureus the gp91phox−/− mice were . Remarkably , we found that the 100% lethal dose ( LD100 ) of C . albicans SC5314 was >150-fold lower in gp91phox−/− mice vs . wild-type controls ( <103 vs . 1 . 5×105 ) . The LD100 of S . aureus LAC was 2-fold lower in gp91phox−/− mice vs . wild type controls ( 5×106 vs . 107 ) . Subsequently , gp91phox−/− and wild type mice were vaccinated with rAls3p-N plus Al ( OH ) 3 or Al ( OH ) 3 alone . CD4+ T cells from vaccinated or control wild type donor mice were adoptively transferred into gp91phox−/− recipient mice , and visa versa . As well , some mice were vaccinated and infected without undergoing adoptive transfer as positive ( wild type ) and negative ( gp91phox−/− ) controls for vaccine efficacy . The vaccine did not improve the survival of gp91phox−/− mice infected with either organism ( Fig . 3A ) . While CD4+ T lymphocytes from vaccinated , gp91phox−/− donor mice improved the survival of wild type recipient mice , CD4+ T cells from vaccinated , wild-type donor mice failed to improve the survival of gp91phox−/− recipient mice ( Fig . 3B ) . The need for downstream functional phagocytes to mediate vaccine efficacy suggested that Th17 cells , which are known to act by recruiting phagocytes to the sites of infection [25] , [26] , might play a role . To determine the requirement for IL-17 and Th17 cells in mediating vaccine efficacy , we vaccinated mice deficient in IL-17A , or their wild type congenic control mice . IL-17A-deficiency abrogated vaccine-mediated efficacy ( Fig . 4A ) . Of note , in contrast to IFN-γ deficiency , IL-17A deficiency did not exacerbate the severity of infection in unvaccinated mice ( comparing survival of unvaccinated , deficient vs . wild type mice ) . To determine if CD4+ T cells were the primary source of IL-17A in mediating vaccine efficacy , CD4+ T cells from vaccinated or control mice were cross-adoptively transferred into recipient mice ( IL-17A-deficient donor cells transferred to wild type recipient mice; wild type donor cells transferred to IL-17A-deficient recipient mice ) . We also repeated the survival study in wild type and IL-17A deficient mice that did not undergo adoptive transfer to serve as positive and negative controls for the adoptive transfer study . Mice were infected the day after adoptive transfer . Once again , the vaccine improved the survival of the positive control wild type mice but not the negative control IL-17A deficient mice ( Fig . 4B ) . Adoptive transfer of CD4+ cells from vaccinated wild type donor mice improved the survival of IL-17A-deficient recipient mice ( Fig . 4B ) . In contrast , transfer of CD4+ T cells from vaccinated IL-17A-deficient donor mice to wild type recipient mice failed to improve survival ( Fig . 4C ) , confirming that CD4+ T cell derived IL-17A was necessary to mediate vaccine efficacy . To define the populations of cells induced by vaccination , spleens and lymph nodes were harvested from vaccinated and control mice two weeks following the boost . The cells were stimulated ex vivo for 5 days with rAls3p-N . Analysis of supernatants confirmed that cells from vaccinated mice produced significantly more IFN-γ and IL-17 , as well as the neutrophil-acting chemokines , KC and MIP-1α , than did cells from control mice ( Fig . 5A ) . IL-4 levels were not detectable in any supernatant from control cells; levels were detectable at low levels ( < 2 pg/ml ) in supernatants from 4 of the 8 mice in the vaccinated group . However , IL-10 and IL-13 levels were higher in supernatants from vaccinated than control mice . Levels of TGF-β and IL-6 were low and not significantly different in supernatants from vaccinated or control mice . Supernatants from stimulated , immune cells markedly enhanced phagocytic killing of C . albicans and S . aureus ex vivo , compared to supernatant from control cells ( Fig . 5B ) . Intracellular cytokine analysis of the stimulated cells demonstrated that vaccination resulted in increased frequencies of Th1 ( CD4+IFN-γ+ ) , Th17 ( CD4+IL-17+ ) , and Th1/17 ( CD4+IFN-γ+IL-17+ ) cells in draining lymph nodes lymphocytes ( Fig . 6 and Figs . S2 and S3 ) compared to the frequencies in unvaccinated mice . Murine CD4+CCR6- cells were enriched for the Th1 phenotype , and CD4+CCR6+ cells were enriched for the Th17 phenotype . However , a substantial proportion of CCR6+ splenocytes , and particularly CD4+CCR6+ lymphocytes , were Th1/17 ( IFNγ+IL-17+ ) cells . The Th1/17 phenotype was predominantly found in CD4+CCR6+ cells , not in the CD4+CCR6- cells . To confirm the in vivo biological relevance of the ex vivo lymphocyte phenotypes , vaccinated or control mice were infected via the tail vein with C . albicans or S . aureus 2 weeks following the boost . At day 4 post-infection ( the day before control mice were anticipated to begin dying ) , burden of infection and cytokine levels in homogenates of individually marked kidneys ( primary target organ ) were determined . Levels of myeloperoxidase ( MPO ) , which is constitutively expressed at the protein level in neutrophils and has been extensively used in previous studies to quantify neutrophil influx into tissues during infection and inflammation [27]–[31] , were also measured . Vaccination resulted in a ∼10-fold reduction in kidney fungal burden and ∼5-fold reduction in kidney bacterial burden ( Fig . 7A ) . MPO levels were increased in vaccinated mice relative to control mice infected with either organism ( Fig . 7B ) . A recent study reported a 95% correlation between organ fungal burden and neutrophil influx in mice infected with different strains of C . albicans or C . dubliniensis [32] . Therefore , any enhanced neutrophil influx resulting from vaccination could be offset by the diminished stimulus for neutrophil influx caused by reduced fungal burden in the vaccinated mice . To isolate the impact on MPO levels of vaccination , and not severity of infection , we adjusted absolute MPO levels in individually marked organs for the fungal or bacterial burden in those individual organs . Vaccination resulted in a marked increase in neutrophil influx relative to the infectious burden of organism in the tissues ( Fig . 7B ) . By histopathology , the inflammatory infiltrate was predominantly neutrophilic , with scattered foci of macrophages . Semi-quantitative scoring of histopathology sections by a blinded pathologist to estimate neutrophil influx into tissues was concordant with the quantitative MPO levels . Concordant with ex vivo cytokine measurements , absolute levels of IFN-γ , IL-17 , and the neutrophil-acting CXC chemokine , KC , were higher in the kidneys of vaccinated versus control mice ( p<0 . 05 for all comparisons of vaccinated vs . control levels for all three cytokines , in mice infected with C . albicans or S . aureus ) . After adjusting for infectious burden in individual organs , vaccination markedly increased cytokine levels relative to infectious burden ( Fig . 7C ) . Histopathology confirmed a marked increase in organism burden in the vaccinated mice versus control mice infected with either organism ( Fig . 8 ) . Numerous microabscesses with hyphal and pseudohyphal elements were scattered throughout the kidneys of control mice infected with C . albicans . Microabscesses were also found in the kidneys of vaccinated mice , but most of the abscesses had no fungal elements visible , and those few abscesses with fungal elements contained blastospores or small hyphal fragments . Control mice infected with S . aureus had large renal abscesses with numerous gram positive cocci on Gram stain . Vaccinated mice also had renal abscesses with extensive neutrophil influx , but in most abscesses fewer staphylococcal organisms were seen on Gram stain in vaccinated than control mice ( Fig . 8 ) .
One hypothesis regarding the failure to date to develop an effective vaccine against S . aureus or Candida has been the need to simultaneously disrupt multiple virulence factors for such complex pathogens , whereas most vaccines to date have targeted only one virulence factor [33] , [34] . However , we have previously confirmed that humoral immunity is neither necessary nor sufficient for rAls3p-N vaccine-induced protection against either organism [6] , [7] . Furthermore , homozygous disruption of ALS3 in C . albicans does not result in a loss of pathogenicity in vivo in mice , so the protection mediated by the rAls3p-N vaccine is not the result of abrogation of Als3p virulence functions . The current study confirms that vaccination can be effective by targeting the organism for destruction by increasing the quantity and microbicidal functions of innate phagocytic effectors at the site of infection , irrespective of affecting virulence functions in the organism . Therefore , potential vaccine antigens need not be restricted to microbial virulence factors , and can be expanded to include any target antigen which results in a potent Th1 and/or Th17 immune response against the organism . These data are concordant with the established role of Th17 cells in mediating protection following immunization of mice against Mycobacterium tuberculosis , Helicobacter pylori and Pseudomonas aeruginosa [31] , [35] , [36] . In unvaccinated animals , deficiency in IFN-γ but not IL-17A exacerbated the severity of iv infection caused by both S . aureus and C . albicans . These results are concordant with recent studies demonstrating that IL-17-deficient mice were not more susceptible to bloodstream infection caused by S . aureus [37] or invasive gastric infection caused by C . albicans [38] . Furthermore , a recent study reported that abrogation of the dectin-2 receptor blocked Th17 induction by C . albicans in mice , but despite the lack of a Th17 response did not affect the ability of mice to clear fungus from tissue during systemic infection [39] . Collectively , these results indicate that Th17 cells/IL-17A are not necessary for normal murine host defense against disseminated candidiasis . In contrast , IL-17 has been shown to be critical for host defense against cutaneous and oropharyngeal infections caused by S . aureus [37] and C . albicans [40] , respectively . Furthermore , our results are discordant with the previous finding that IL-17 receptor-deficiency moderately exacerbated the severity of bloodstream infection caused by C . albicans [41] . The lack of a requirement for IL-17A to mediate normal host defense against disseminated candidiasis likely reflects the ability of IL-17F , which also activates the common IL-17 receptor , to complement an IL-17A deficiency . Differences in C . albicans infecting strain and mouse host strain may also account for differences between our study and the prior . However , a critical point is that IL-17F could not compensate for the requirement for IL-17A in mediating protective , vaccine-induced , adaptive immunity , since IL-17A deficiency abrogated vaccine efficacy . We confirmed that the rAls3p-N vaccine specifically primed splenic and lymph node lymphocytes to produce high levels of both IFN-γ and IL-17 , as well as the neutrophil chemokines , KC and MIP-1α ( the latter of which is chemotactic for both mononuclear cells and neutrophils [42]–[47] ) . The predominant IFN-γ expression in lymph nodes was found in CCR6- Th1 cells which did not produce IL-17 ( CD4+CCR6−IFN-γ+IL-17− ) , and the predominant IL-17 expression in lymph nodes was found in CCR6+ cells . However , we also found substantial numbers of Th1/17 cells , which met or exceeded the frequency of Th17 cells , in the CD4+CCR6+ fraction . The Th1/17 cells were found virtually exclusively in the CCR6+ fraction , and none were found in the CD4+CCR6- fraction . Recent studies have indicated that yeast mannosylated proteins prime Th17 cells via activation of the mannose receptor [48] , and that O-linked mannoproteins can activate IFN-γ production via ligation of TLR4 [49] . Since the rAls3p-N protein has O-linked yeast high mannose groups , co-ligation of the mannose receptor and TLR4 on antigen presenting cells may enable induction of Th1 , Th17 , and Th1/17 cells . The role of specific antigen presenting cells in priming lymphocytes for Th1 , Th17 , or dual Th1/17 responses is currently under investigation . We found variations in the total number of surviving mice from experiment to experiment , ranging from as high as 87% to as low as 12 . 5% . Variations in outcome are most likely accounted for by variations in infectious inoculum and infecting strain . Our challenge model , using the standard SC5314 clinical isolate of C . albicans , is extremely rigorous , and is considerably more rigorous than challenge with other clinical strains of C . albicans [6] , [50] , [51] , as evidenced by the superior efficacy seen in the current study with another clinical bloodstream isolate of C . albicans ( 15563 ) . We have previously shown that mice infected with the inocula of SC5314 used in these experiments die of overwhelming septic shock [52] . Candidal septic shock causes >50% mortality in humans despite treatment with antifungal therapy [4] . Hence , achievement of survival approaching 50% by vaccination alone is felt to reflect meaningful protection . Furthermore , the experiment in which 12 . 5% survival was seen in the vaccinated arm was an adoptive transfer experiment , in which immune cells from wild type mice were transferred into IL-17A-/- recipient mice . Thus , while IL-17A production from CD4+ immune T cells can transfer protection , production of IL-17A by other cell types may be required to achieve maximal protection . Specifically , we previously found that immune CD8+ T cells could transfer protection against S . aureus [7] , and macrophages or dendritic cells can produce pro-inflammatory cytokines such as IFN-γ , suggesting that these cell types may play an adjunctive role and be required for full vaccine-mediated protection . We previously reported that cyclophosphamide-induced neutropenia did not completely abrogate vaccine-induced protection during subsequent disseminated candidiasis [53] . In contrast , in the current study , we did find total abrogation of protection against both candidal strains and against S . aureus . The prior study used a different but related vaccine immunogen , rAls1p-N , instead of rAls3p-N . As well , the prior study used Complete Freund's Adjuvant ( CFA ) , not Al ( OH ) 3 . The greater efficacy of the former adjuvant may account for the residual efficacy found in neutropenic mice in the former study . S . aureus and C . albicans express adhesins on their cell surface which possess similar three dimensional shapes [54] and which bind to similar endovascular surfaces ( e . g . endothelial cells and subendothelial matrix proteins ) and medically relevant plastics [54] , [55] . Given these similar virulence mechanisms , it is not surprising that the organisms also infect patients with similar risk factors , including post-operative and trauma patients , patients with central venous catheters , patients on hemodialysis , and patients with compromised phagocytic host defense mechanisms [4] , [56] , [57] . Finally , our data demonstrate that the host defends itself against both infections by similar mechanisms , and that adaptive immunity to both organisms required CD4+ T cell production of both IFN-γ and IL-17A . In summary , the rAls3p-N vaccine improved outcomes in mouse models of iv S . aureus and C . albicans infection by inducing upstream , pro-inflammatory , Th1 , Th17 , and Th1/17 lymphocytes , which enhanced recruitment and activation of neutrophils in infected tissues , thereby reducing tissue infectious burden . Thus , vaccination showed a potential to protect against both infections by targeting the microbes for enhanced destruction by innate effector cells , irrespective of neutralization of microbial virulence factors . Therefore , potential vaccine antigens need not be restricted to microbial virulence factors , and can be expanded to include any target antigen which results in a potent Th1 and/or Th17 immune response against the organisms .
C . albicans SC5314 was supplied by W . Fonzi ( Georgetown University ) , and S . aureus LAC , a USA300 MRSA clinical isolate , was provided by Frank Deleo ( NIAID/NIH ) . C . albicans 15563 is a clinical bloodstream isolate from a patient at Harbor-UCLA Medical Center which is also virulent in our murine model [50] . Candida was serially passaged three times in yeast peptone dextrose broth ( Difco ) at room temperature prior to infection . S . aureus was grown overnight at 37°C in BHI broth , and then passaged for 4 hours at 37°C in fresh BHI broth . Female Balb/c or C57BL/6 mice were obtained from Taconic Farms ( Bethesda , MD ) . Congenic IL-17A deficient mice on a Balb/c background were obtained from Y . Iwakura ( University of Tokyo ) [58] . Vaccinated mice were infected via the tail vein with the appropriate inocula of C . albicans blastospores or S . aureus organisms in PBS , as previously described [7] , [52] . In some experiments , mice were made neutropenic by treatment with 230 mg/kg cyclophosphamide 2 days prior to infection , a regimen which results in profound neutropenia for approximately one week [59] , [60] . All procedures involving mice were approved by the Los Angeles Biomedical Research Institute animal use and care committee , following the National Institutes of Health guidelines for animal housing and care . rAls3p-N ( amino acids 17 to 432 of Als3p ) was produced in Saccharomyces cerevisiae and purified by Ni-NTA matrix affinity purification as previously described [61] . Mice were immunized by subcutaneous ( SQ ) injection of 300 µg of rAls3p-N in 0 . 1% Al ( OH ) 3 ( Alhydrogel , Brenntag Biosector , Frederikssund , Denmark ) in PBS . Control mice received adjuvant alone on the same schedule . Some mice were boosted at 21 days . Mice were infected two weeks following the boost . Serum and splenic lymphocytes were harvested from vaccinated or control mice , as we have previously described [62] . Lymph node lymphocytes were harvested from cervical and axillary lymph nodes , based on pilot studies with Evans Blue dye lymph node mapping demonstrating that SQ vaccination at the base of the neck drained primarily to these lymph nodes . For adoptive transfers , splenic and lymph node lymphocytes were pooled . CD4+ T lymphocytes were purified by use of the IMag system ( BD Pharmingen ) , as we have described [6] , [7] . Purified lymphocytes ( 5×106 per mouse ) were administered iv to congenic , unvaccinated recipient mice . Transferred cell populations were ≥95% pure by flow cytometric analysis . Mice were infected via the tail-vein with C . albicans SC5314 24 h after lymphocyte adoptive transfer . Intracellular cytokines from lymphocytes were analyzed based on a modification of our previously described method [62] . In brief , cervical and axillary lymph nodes and spleens were dissected from vaccinated or control mice and passed through 70 µm filters . Cells were stimulated for 5 days with rAls3p-N ( 12 . 5 µg/ml ) in complete media ( RPMI 1640 , 50 U/ml penicillin , 50 µg/ml streptomycin , 2 mM L-glutamine , 10% FBS , 5 µM 2-ME ) in 96 well plates . PMA ( 50 ng/ml ) , ionomycin ( 1 µM ) , and monensin ( 10 µg/ml ) were added during the final 6 hours of culture . Supernatant was harvested prior to adding monensin for analysis of cytokine content using Cytometric Bead Array Flex kits ( BD Pharmingen , La Jolla , CA ) or ELISA ( for IL-6 , TGF-β , and IL-13 ) , per the manufacturer's instructions . Cells were stained on ice with PerCP-anti-CD4 and Alexa647-anti-CCR6 ( BD Pharmingen , San Diego ) , or their isotype control antibodies . The cells were fixed and permeabilized as previously described [62] . Intracellular cytokines were stained with rat FITC-anti-mouse IFN-γ and PE-anti-IL-17 , or their isotype controls ( BD Pharmingen ) . Four-color flow cytometry was performed on a Becton-Dickinson FACScan instrument calibrated with CaliBRITE beads ( Becton Dickinson , San Jose , CA ) using FACSComp software as per the manufacturer's recommendations . Data for each sample were acquired until 10 , 000 CD4+ lymphocytes were analyzed . Th1 cells were defined as CD4+IFNγ+IL-17− , Th17 cells defined as CD4+IFN-γ−IL-17+ , and Th1/17 cells defined CD4+IFNγ+IL-17+ . The killing assay for both C . albicans and S . aureus was modified based on our well-described assay [59] , [60] . In brief , RAW murine macrophage cells or murine neutrophil cells were grown in DMEM plus 10% fetal bovine serum . Fresh murine neutrophils were harvested by dextran sedimentation of whole , heparanized blood , followed by centrifugation over Ficoll Hypaque at 500 g for 10 minutes . The RAW cells or neutrophils were added into 24 well plates , the media in the wells was aspirated and the RAW cells or fresh neutrophils were cultured for 4 hours in 10% conditioned media ( from vaccinated or control splenic and lymph node lymphocytes exposed to rAls3p-N for 5 days ) plus 90% complete media ( RPMI + 10% FBS ) . The conditioned media was then aspirated , and the microorganisms added to the wells in fresh DMEM plus 10% fetal bovine serum . Microorganisms were added to the wells at a ratio of 20∶1 RAW cells to C . albicans , 5∶1 RAW cells to S . aureus , or 10∶1 fresh neutrophils to C . albicans or S . aureus . Media for the wells containing S . aureus contained no antibiotics . The cells were incubated at 37°C for 1 h , at which point 4% blood heart infusion ( BHI ) agar was directly added to the wells . Plates were incubated overnight at 37°C and colony forming units ( CFUs ) counted in each well . Killing was defined as the percent reduction CFUs in wells containing co-cultures of phagocytes cells and microorganisms compared to wells just containing microorganisms . On day 4 post-infection , kidneys ( primary target organ ) were harvested and homogenized in saline with protease inhibitors ( pepstatin , leupeptin , and PMFS ) . For determination of infectious burden , organ homogenates were quantitatively cultured on Sabourad dextrose agar for C . albicans or tryptic soy agar for S . aureus . Whole organ cytokines were analyzed from kidney homogenates by ELISA ( R&D Systems ) or Cytometric Bead Array Flex kit for KC ( BD Pharmingen , La Jolla , CA ) , per the manufacturer's instructions . MPO levels were determined by ELISA ( Hycult Biotechnology , Uden , Netherlands ) of whole organ homogenates . For histopathology , organs were fixed in zinc-buffered formalin , embedded in paraffin , sectioned , and stained with PAS for fungi and H&E and Gram stain for bacteria . The non-parametric Log Rank test was utilized to determine differences in survival times . The Wilcoxon Rank test was used to compare cytokines , MPO levels , and organ burden across groups . P<0 . 05 was considered significant . | The bacterium Staphylococcus aureus and the fungus Candida are the second and third leading cause of bloodstream infections in hospitalized patients . A vaccine to prevent such infections would be of enormous public health benefit . The leading hypothesis to explain why vaccines have not been successfully developed against these infections is that the microbes causing the infections are highly complex , and use multiple weapons ( so-called “virulence factors” ) to cause disease in humans . Therefore , a vaccine targeting either infection would have to neutralize many of these virulence factors at the same time . We have been developing a vaccine that simultaneously targets both types of infections . Our vaccine is based on a single virulence factor used by Candida , which has a similar shape to virulence factors used by S . aureus . In the current study , we report that our vaccine induces specialized cells in the immune system to more effectively call in reinforcements to kill the organisms . These data demonstrate that vaccines against both organisms can be developed even if they do not work by neutralizing multiple virulence factors , and therefore open the door to a far wider array of vaccine types against both infections . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"infectious",
"diseases/bacterial",
"infections",
"infectious",
"diseases/fungal",
"infections",
"immunology/immunity",
"to",
"infections"
] | 2009 | Th1-Th17 Cells Mediate Protective Adaptive Immunity against Staphylococcus aureus and Candida albicans Infection in Mice |
Since its initiation in 1995 , the African Program for Onchocerciasis Control ( APOC ) has had a substantial impact on the prevalence and burden of onchocerciasis through annual ivermectin mass treatment . Ivermectin is a broad-spectrum anti-parasitic agent that also has an impact on other co-endemic parasitic infections . In this study , we roughly assessed the additional impact of APOC activities on the burden of the most important off-target infections: soil-transmitted helminthiases ( STH; ascariasis , trichuriasis , hookworm , and strongyloidiasis ) , lymphatic filariasis ( LF ) , and scabies . Based on a literature review , we formulated assumptions about the impact of ivermectin treatment on the disease burden of these off-target infections . Using data on the number of ivermectin treatments in APOC regions and the latest estimates of the burden of disease , we then calculated the impact of APOC activities on off-target infections in terms of disability-adjusted life years ( DALYs ) averted . We conservatively estimated that between 1995 and 2010 , annual ivermectin mass treatment has cumulatively averted about 500 thousand DALYs from co-endemic STH infections , LF , and scabies . This impact comprised approximately an additional 5 . 5% relative to the total burden averted from onchocerciasis ( 8 . 9 million DALYs ) and indicates that the overall cost-effectiveness of APOC is even higher than previously reported .
The African Program for Onchocerciasis Control ( APOC ) is an international program aimed at controlling the disease burden of human onchocerciasis ( river blindness ) in sub-Saharan Africa ( SSA ) , and elimination of infection where possible , using mass drug treatment ( MDA ) [1 , 2] . Since its launch in 1995 , APOC and partnering beneficiary countries have scaled up their control activities geographically to at least cover all meso- and hyperendemic areas , averting 8 . 9 million disability-adjusted life years ( DALYs ) through 2010 , and eventually aiming to treat over 90 million people annually in 16 African countries by 2015 , protecting a population at risk of onchocerciasis of 118 million [3 , 4] . The drug used for mass treatment of onchocerciasis , ivermectin , is distributed and administered in a single dose of 150–200 μg/kg of body weight annually . Chronically ill people , pregnant ( or lactating ) women , and children under five are excluded from treatment with ivermectin [1] . Ivermectin is known to be effective against various infectious diseases other than onchocerciasis , the most important being soil-transmitted helminth ( STH ) infections , lymphatic filariasis ( LF ) , and epidermal parasitic skin diseases ( EPSDs ) such as scabies [5–11] . In APOC countries , the prevalence of STH infections in school-age children ranges between 20% and 50% [12] . LF is endemic in all APOC countries with an estimated overall prevalence of 6–9% [10] , and local prevalences typically ranging between 0–40% [13] . Despite the lack of comprehensive epidemiological data , it is known that EPSDs are prevalent across SSA and that the associated morbidity is significant in regions of high poverty [9 , 14 , 15] . Together , these infections are responsible for a considerable burden of disease [14 , 15] . Therefore , annual mass treatment with ivermectin is expected to have an additional health impact by averting part of the burden related to these off-target infections [8 , 16–19] . Although these additional beneficial effects of ivermectin are being used to sensitise communities to participate in MDA , up till now , the off-target health impact has not been quantified and its importance remains unknown . In this study , we quantified the health impact of APOC activities through 2010 on the burden of STH ( ascariasis , trichuriasis , hookworm , and strongyloidiasis ) , LF , and scabies . We reviewed the literature to retrieve field studies examining the effect of ivermectin treatment on off-target infections and formulated assumptions about the impact of ivermectin mass treatment on the associated burden of disease . Next , we retrieved estimates of the disease burden of candidate off-target diseases from the Global Burden of Disease ( GBD ) 2010 Study [14] . By combining this information with data on the number of ivermectin treatments given through 2010 ( recorded by APOC ) , we roughly estimated the number of DALYs due to off-target infections averted by APOC .
We first performed a systematic PubMed search to determine efficacy of ivermectin mono-treatment against off-target infection , defined as the ability to provide a clinically measurable and preferably beneficial effect . We used the key term “ivermectin” in combination with any of the following: “efficacy” , “mass treatment” , “morbidity control” . Searches were made without time limitations . If available , meta-analysis studies evaluating the efficacy of ivermectin against a specific disease were used . If meta-analysis studies were not available , clinical studies reporting the efficacy of treatment were selected if: ( 1 ) the treatment regime concerned a single dose of about 150–200 μg/kg of body weight , and ( 2 ) the efficacy was evaluated up to one month for STH infections and up to a year for filarial infections and EPSDs . We considered a month to be the threshold duration of the immediate effect of ivermectin on STH infections . For LF and EPSDs , longer periods were considered due to a lack of studies evaluating the efficacy of ivermectin as soon as one month after administration . Only studies reporting their results in terms of the following criteria were considered: ( 1 ) percent of patients cured and/or percent egg reduction for STH infections , ( 2 ) percent microfilaria ( mf ) reduction ( microfilaricidal efficacy ) and/or the percent reduction in female fecundity ( embryostatic efficacy ) in LF , and ( 3 ) percent of patients cured for EPSDs . For some EPSDs , clinical studies describing single cases were considered due to the rareness of their incidence . If repeated doses were given , it was noted . Based on the results of the literature review , we formulated assumptions about the effect of ivermectin mono-treatment on the burden of ( untreated ) off-target infections , in terms of reduction in DALYs lost ( Box 1 ) . Assumptions were formulated while considering the following factors: the direct effect of ivermectin treatment on infection levels in individuals , the clinical manifestations of each disease , the short and long term effects of mass treatment on incidence and prevalence of morbidity , and the patterns of post-treatment re-infection . The effect of ivermectin was expressed as parameter βx ( range 0–1 ) , which represents the average reduction in the burden of disease x in DALYs lost over a period of six years achieved thanks to mass treatment with ivermectin . The six-year period was based on APOC data on population coverage of ivermectin mass treatment , which suggest that most of the population in APOC areas has been subject to at least six rounds of mass treatment between 1995 and 2010 [3] . For infections such as STH and scabies , in which morbidity is highly correlated with intensity of infection ( parasite load ) , and treatment only influences transmission to a small extent , we assumed that the impact of treatment is the same each year . For LF , repeated mass treatment rounds are expected to have an increasingly higher impact on the disease burden , through the effects of mass treatment on transmission and prevention of further exposure to infection that would lead to chronic disability ( e . g . lymphedema ) . We assumed that for LF , parameter βx represents the average health impact in all areas subject to varying periods of ivermectin mass treatment . From the GBD 2010 study [14] , we derived country-specific estimates of the burden per capita ( DALYs lost per 100 , 000 persons ) for ascariasis , trichuriasis , hookworm , LF , and EPSDs ( scabies only ) in 1995 , 2000 , 2005 , and 2010 ( extracted from the online GBD data visualization tool [20] ) . For the years in between , we assumed that the disease burden of these off-target infections followed a trend consistent with exponential interpolation of the available estimates . For some countries covered by APOC , the GBD 2010 study reports a decline in the burden of off-target diseases between 1995 and 2010 . In the current study , we assume that this decline is not due to APOC or other control activities , at worst leading to an underestimation of the health impact of ivermectin mass treatment on off-target diseases ( i . e . because the counterfactual burden without these activities would be even higher than reported by GBD 2010 ) . Though the group of EPSDs consists of several infections such as scabies , tungiasis ( sand fleas ) , pediculosis ( lice ) , and several other infections [9] , for the purpose of this study , we only considered scabies , as burden estimates have been made only for this particular infection so far . Further , the GBD 2010 study does not provide estimates for the burden of strongyloidiasis , an STH , even though its prevalence in SSA is probably considerable ( although of unknown size ) . We assumed that the burden due to strongyloidiasis amounts to 1/5 of the total burden caused by the three major STH infections ( ascariasis , trichuriasis , and hookworm ) . This assumption was based on the estimate that the prevalence of the three major STH infections in SSA ranges between 20–50% [12] , and a large cross-sectional study in rural Ghana which reported a prevalence of strongyloidiasis of 11 . 6% [21] . These figures suggest that the prevalence and presumably the burden of strongyloidiasis may be up to five times lower than the prevalence and burden of the other three more common STH infections . Given the uncertainty regarding the assumption , we varied it in the sensitivity analysis ( details below ) . We calculated the disease burden averted , Maiyx ( in DALYs ) , for each selected off-target infection x and summed it over the sixteen countries in which APOC has been active between 1995 and 2010 using the following formula , where i represents a specific APOC country and y represents the year of mass treatment: Max=∑i=116∑y=19952010Maiyx=∑i=116∑y=19952010βxTiyhix1−pMcx−pMwx1−pc−pwMiyx In this formula , Mixy reflects the annual burden per capita due to infection x ( as reported by the GBD 2010 study ) . This figure was adjusted to represent the burden per capita in population eligible for mass treatment with ivermectin in APOC areas by adjusting for over- or underrepresentation of the disease burden in children under five and pregnant or lactating women who are not eligible for ivermectin ( 1−pMcx−pMwx1−pc−pw is the fraction proportion of burden in ivermectin-eligible populationproportion of population eligible for ivermectin , where pMcx + pMwx is the proportion of the disease burden in children under five , and pc + pw is the proportion of children under five and pregnant or lactating women in the population ) . We further adjusted for clustering of disease burden in APOC regions compared to other country regions ( hix ) . The adjusted burden per capita ( hix1−pMcx−pMwx1−pc−pwMixy ) was then multiplied by the number of treated individuals Tiy for any given year y ( extracted from APOC records ) , yielding the potential disease burden in treated people . Assumptions about the values of aforementioned parameters can be found in Table 1 , along with the associated literature references [3 , 11 , 12 , 14 , 20 , 22–27] . The potential disease burden in treated people was multiplied with the infection-specific ivermectin efficacy βx , yielding the estimated averted disease burden related to infection x in country i for year y . Results were then summed over years and countries , resulting in a total estimated number of DALYs averted ( Max ) , related to infection x . A univariate sensitivity analysis was performed to assess how the main result ( number of DALYs averted related to off-target infections ) changed when parameter βx was increased or decreased by 20% . The value of 20% was chosen because larger increases or decreases of the effectiveness of ivermectin against off-target infections are highly unlikely to occur in field settings . Likewise , the assigned value of hix was varied by ±20% for those countries where hix was initially set to 1 . 5 ( see Table 1 ) . In addition , prevalence and burden of STH infections or EPSDs were assumed to either be overrepresented in APOC regions for all or none of the countries covered by APOC . The impact of the assumption regarding burden due to strongyloidiasis ( 1/5 of the total burden due to other STH infections ) was examined by halving or doubling the assumed proportion . Finally , we also performed a multivariate sensitivity analysis by simultaneously increasing or decreasing all assumed βx parameters by 20% as an extreme scenario .
We assumed that each year , mass treatment would avert some fraction βx of the potential disease burden in treated communities . Based on literature , this fraction was assumed to be 0 . 5 for ascariasis , trichuriasis , strongyloidiasis and EPSDs , 0 . 2 for hookworm infections , and 0 . 1 for LF ( Box 1 ) . We estimated that without APOC ( counterfactual situation assuming that there is no mass treatment with ivermectin by APOC ) , the potential disease burden of STH infections , lymphatic filariasis , and EPSD in individuals otherwise treated with ivermectin would amount to a cumulative burden of 1 . 7 million DALYs between 1995 and 2010 . Of these , 493 thousand DALYs were averted by APOC through ivermectin mass treatment ( Table 2 ) . Most of the DALYs averted by APOC were related to ascariasis ( 162 thousand ) and scabies ( 116 thousand ) , followed by LF ( 71 thousand ) , strongyloidiasis ( 67 thousand ) , and hookworm infection ( 61 thousand ) . Only a small part of the burden averted by APOC was related to trichuriasis ( 17 thousand ) . Nigeria contributed 53% of the total averted number of DALYs related to off-target infections ( 260 thousand of 493 thousand ) , followed by the DRC ( 62 thousand or 13% ) and Cameroon ( 59 thousand or 12% ) . Fig 1 shows the results of the sensitivity analyses . Changing individual βx parameters by 20% resulted in estimates very similar to the main estimate . Obviously , increasing or decreasing all βx parameters simultaneously by 20% resulted in ±20% deviations from the main estimate of 493 thousand DALYs averted . Assumed no clustering of STH and EPSDs in APOC areas ( hix = 1 . 0 ) or clustering of all infections but LF in all countries ( hix = 1 . 5 ) resulted in 16% lower and 19% higher estimates of total DALYs averted , respectively .
The impact of APOC on off-target NTDs has previously been discussed and considered to be important , but difficult to quantify . We estimated that if APOC would not have been there , STH infections , strongyloidiasis , and scabies would have caused a cumulative burden of 1 . 7 million DALYs lost between 1995 and 2010 in individuals who would otherwise have been treated with ivermectin . We roughly estimated that of these 1 . 7 million DALYs , mass treatment with ivermectin has averted 500 thousand DALYs . This means that apart from the impact of APOC on the burden of onchocerciasis ( 8 . 9 million DALYs averted ) , there has been an additional 5 . 5% health impact through the effect of ivermectin mass treatment on off-target NTDs . This indicates that the cost-effectiveness of APOC is even somewhat higher than previously estimated . The estimate of 500 thousand additionally averted DALYs was based on a simple approach that included assumptions about the impact of mass treatment on the burden of selected off-target infections endemic in APOC countries . Because we considered a period of six years for estimating the effects of ivermectin mass treatment ( i . e . the minimum duration of most APOC programs ) , our approach may underestimate the averted burden in countries where ivermectin mass treatments have taken place for over six years ( i . e . where effects on transmission may be larger ) . Also , we did not take into account the protective impact of ivermectin mass treatment due to a reduced transmission to children under five years of age and pregnant women ( who receive no ivermectin ) . Gutman et al . show that the prevalence of some STH infections was significantly lower in pre-school children living in treated communities compared to pre-school children living in non-treated communities [19] . Further , we assumed that APOC interventions have not been accounted for in the burden estimates for off-target NTDs provided by the GBD 2010 study , meaning that at worst ( if GBD 2010 does account for APOC ) , ivermectin mass treatment has had a larger health impact than we estimate here . Also , ivermectin mass treatment probably has an effect on the burden of relatively rare or minor infections that were excluded from our analysis , such as enterobiasis , loiasis , streptocerciasis , serous cavity filariasis , and EPSDs other than scabies . Furthermore , ivermectin mass treatment also has a–yet to be quantified–effect on malaria transmission through the endectocidal effects of ivermectin on Anopheles vectors [64] . On the other hand , we did not consider the burden of severe adverse effects of ivermectin treatment related to loiasis [65 , 66] , which is endemic in parts of the APOC region [67] . Overall , if anything , our results underestimate the true impact of APOC activities on off-target infections . Our estimates of the impact of APOC on the burden of off-target diseases could be further refined with more sophisticated approaches , such as mathematical modeling . For some of the off-target infections , mathematical models have already been developed , such as for transmission and morbidity due ascariasis and transmission of lymphatic filariasis [22 , 68] . Epidemiological data and understanding of the mechanisms through which parasitic infections cause morbidity in the human host are needed to develop similar models for other parasitic infection , and update currently existing models . However , estimates made with such models are only usefully accurate if they are based on good information about the distribution of worms in host populations . Since such data are not yet widely available and the development of mathematical models is time-consuming and expensive , obtaining more precise estimates of the ( averted ) burden of off-target infections remains a challenge . We ignored that in 16 countries , ivermectin mass treatment was planned to be combined with albendazole to target both onchocerciasis and LF from 2000 onwards [8] . However , 11 countries had not provided any LF treatment by 2010 ( Angola , Burundi , Chad , Congo , DRC , Equatorial Guinea , Liberia , Sudan , South Sudan ) , and two countries had only implemented 1–2 treatment rounds reaching <6% of the target population ( Ethiopia and Central African Republic ) [69] . Only five countries ( Cameroon , Malawi , Nigeria , Uganda , United Republic of Tanzania ) had more mature LF elimination programmes that supplied 3–11 treatment rounds through 2010 , reaching 25%–83% of the LF target population . We lack information on whether LF treatments took place in areas with LF only or co-endemic areas , but countries might have started in areas where mass treatment already took place for onchocerciasis . For these countries ( especially Nigeria where many people live in co-endemic areas ) it is therefore perhaps not justified to fully attribute the estimated effects on off-target diseases to APOC alone . It would be more useful to estimate the overall effect of repeated mass treatments on all target and off-target diseases; such estimates would be larger than the figures presented here , thanks to the addition of albendazole . The so-called neglected tropical diseases ( consisting of STH , filariases , EPSDs , and several other infections ) are the most common conditions affecting the poorest 500 million people living in SSA [10 , 15] . The infections covered in our analysis have been estimated to be responsible for a burden of 8 . 3 million DALYs lost in SSA in 2010 [14] . Compared to this , our estimate of the health impact of ivermectin mass treatment on off-target diseases is modest ( 493 thousand DALYs averted ) . In order to enhance the impact of mass treatment on LF and STH infections , it would be interesting to consider adding albendazole or other STH-specific drugs to mass treatments ( targeting the appropriate age-groups ) . For instance , in a counterfactual scenario where APOC would have added albendazole to their community-directed treatment strategy from 2000 onwards , an additional 389 thousand DALYs would have been averted by 2010 , assuming that MDA would have had a higher impact on ascariasis , trichuriasis , hookworm , and LF ( i . e . βSTH = 0 . 9 , instead of 0 . 5; βLF = 0 . 3 , instead of 0 . 1 ) . Guidelines for what anthelminthic drugs should be used in different areas have already been formulated [70] . Further , to sustain the off-target health impact after onchocerciasis and LF have been eliminated from APOC target areas , it should be considered to continue mass treatments against STH with albendazole ( possibly combined with ivermectin against trichuriasis [71] ) , either school-based ( mainly covering the burden of ascariasis and trichuriasis in children ) or community-based ( covering the burden of all STH ) . This would also be in line with the London Declaration on Neglected Tropical Diseases [72 , 73] . In conclusion , we roughly estimated that ivermectin mass treatment coordinated by APOC has averted about 500 thousand DALYs related to off-target infections . This health impact constitutes an additional 5 . 5% on top of the impact of APOC on the burden of onchocerciasis , and indicates that the cost-effectiveness of APOC is even higher than previously estimated . To amplify and sustain this additional health impact , control programs could consider adding albendazole to mass treatments , and continue this after elimination of onchocerciasis and lymphatic filariasis . | Onchocerciasis , or river blindness , is an infectious disease caused by the worm Onchocerca volvulus , which is transmitted between humans through the bites of blackflies and causes deforming skin disease , itch , and vision loss . The African Programme for Onchocerciasis Control ( APOC ) aims to control morbidity due to onchocerciasis by implementing mass drug administration ( MDA ) with ivermectin in endemic areas , targeting the whole population except for children under five and pregnant women . Aside from its effect on onchocerciasis , ivermectin also affects other parasitic infections such as lymphatic filariasis , intestinal worm infections , and scabies , which are all significantly co-endemic in areas covered by APOC . In this paper , the researchers roughly estimate the health impact of ivermectin MDA on off-target infections based on the number of dispensed treatments up to 2010 , published estimates of the disease burden of off-target infections , and the expected effect of ivermectin treatment on the burden of these infections ( based on literature review ) . This off-target health impact of APOC constitutes about 500 thousand years worth of healthy years of life ( an additional 5 . 5% on top of the impact of APOC on the burden of onchocerciasis ) and indicates that the cost-effectiveness of APOC is even higher than previously estimated . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [] | 2015 | African Program for Onchocerciasis Control 1995–2010: Impact of Annual Ivermectin Mass Treatment on Off-Target Infectious Diseases |
TcSMUG L products were recently identified as novel mucin-type glycoconjugates restricted to the surface of insect-dwelling epimastigote forms of Trypanosoma cruzi , the etiological agent of Chagas disease . The remarkable conservation of their predicted mature N-terminal region , which is exposed to the extracellular milieu , suggests that TcSMUG L products may be involved in structural and/or functional aspects of the interaction with the insect vector . Here , we investigated the putative roles of TcSMUG L mucins in both in vivo development and ex vivo attachment of epimastigotes to the luminal surface of the digestive tract of Rhodnius prolixus . Our results indicate that the exogenous addition of TcSMUG L N-terminal peptide , but not control T . cruzi mucin peptides , to the infected bloodmeal inhibited the development of parasites in R . prolixus in a dose-dependent manner . Pre-incubation of insect midguts with the TcSMUG L peptide impaired the ex vivo attachment of epimastigotes to the luminal surface epithelium , likely by competing out TcSMUG L binding sites on the luminal surface of the posterior midgut , as revealed by fluorescence microscopy . Together , these observations indicate that TcSMUG L mucins are a determinant of both adhesion of T . cruzi epimastigotes to the posterior midgut epithelial cells of the triatomine , and the infection of the insect vector , R . prolixus .
Described by its discoverer , Carlos Chagas [1] , [2] , as “one of the most injurious tropical illnesses , specially to children in contaminated areas , either in determining a chronic sickly condition in which people become unable to perform vital activities or as an important factor of human degeneration , ” Chagas disease remains a major tropical human disease in much of Latin America , affecting approximately 11 million people . There are 300 , 000 new cases of Chagas disease each year , with approximately 21 , 000 deaths annually [3] . Various triatomine vectors , including Rhodnius , Triatoma and Pastrongylus , are able to acquire and transmit Trypanosoma cruzi , the etiological agent of Chagas disease [4] , [5] . During their development within insects , parasites undergo profound morphological changes , modulating surface molecules to enable interactions with specific insect tissues that are essential for their survival , development and successful transmission to a vertebrate host [6] , [7] . T . cruzi-insect vector interactions begin when the insect feeds on the blood of an infected vertebrate host . Once ingested , most of the bloodstream trypomastigotes differentiate into non-infective epimastigote forms . In the posterior midgut , they repeatedly divide by binary fission and adhere to perimicrovillar membranes ( PMM ) secreted by the underlying midgut epithelial cells [8]–[11] . In the rectum , a proportion of epimastigotes attaches to the rectal cuticle through hydrophobic interactions and transforms into non-replicative infective metacyclic trypomastigotes , which are released together with insect feces and urine during blood feeding [12]–[14] . The entire surface of T . cruzi is covered in glycosylphosphatidylinositol ( GPI ) -anchored mucin molecules that determine parasite protection and establishment of a persistent infection in vertebrate hosts [15] . T . cruzi mucins comprise a large gene family that can be split into two major groups , termed T . cruzi mucin gene family ( TcMUC ) and T . cruzi small mucin-like gene family ( TcSMUG ) , based on sequence comparisons [16] . TcMUC codes for more than 1 , 000 polymorphic products , which are largely co-expressed on the surface of the mammal-dwelling stages [16]–[18] . In addition to their putative immune modulatory role [17] , [19] , one particular TcMUC product termed TSSA ( trypomastigote small surface antigen ) was recently shown to be involved in trypomastigote adhesion to non-macrophagic cells [20] . The second mucin group , TcSMUG , displays significantly less diversity and codes for very small open reading frames . Upon processing of the signal peptide and GPI-anchoring signal , the average predicted molecular mass for the mature apo-mucins would be ∼7 kDa , with Thr representing as much as 50% of the residues . The hydroxyl groups of some of these Thr residues are further derivatized with short O-linked oligosaccharide chains in the Golgi/post-Golgi compartments , which increases the molecular mass of the mature mucins to 35–50 kDa , depending on both the particular TcSMUG product and the parasite isolate [21] , [22] . TcSMUG is composed of two subgroups of genes , named L and S , which display >80% identity on average . Mass spectrometry analyses identified TcSMUG S products as the backbone for the 35/50 kDa mucins ( known as Gp35/50 mucins ) expressed on the surface of insect-dwelling stages [22] . Upon transmission to the mammalian host , Gp35/50 mucins on the surface of metacyclic trypomastigotes bind to non-macrophagic cells in a receptor-mediated manner and induce a bidirectional Ca2+ response , which likely contributes to host-cell invasion [15] . Recent data indicated that TcSMUG L products , though not revealed in the T . cruzi proteomic data sets published so far , constitute a novel mucin-type glycoconjugate restricted to epimastigote forms [22]–[26] . In addition to displaying substantial structural homologies and a common evolutionary origin , comparative analyses highlighted certain differences between TcSMUG L and TcSMUG S products [26] . First , TcSMUG L products , unlike those of TcSMUG S , are not acceptors of sialic acid residues , likely due to the absence of terminal β-Gal residues in the proper configuration . Secondly , and at variance with TcSMUG S products that are expressed at fairly similar levels on every T . cruzi stock , TcSMUG L expression seems quite variable among different parasite isolates . Finally , the remarkable conservation of TcSMUG L deduced products within the predicted mature N-terminal peptide , which does not undergo O-glycosylation , suggest that they are under positive selection against diversification [26] . Because of these features , it has been speculated that structural and/or functional constraints rather than immunological issues limit TcSMUG diversification . In the present work , we investigated the role of TcSMUG L mucins in the attachment of T . cruzi epimastigotes from the Dm28c stock to the midgut epithelium of R . prolixus and the consequent development of the protozoan in the insect vector .
R . prolixus ( Hemiptera: Reduviidae ) were obtained from a longstanding colony reared in the laboratory at 28°C and 60–70% relative humidity [27] where they were fed on chickens weekly and raised as previously described [28] . For the in vivo experiments , the insects were fasted for approximately 15 days and were then fed with infected heat-inactivated citrated human blood using an artificial apparatus similar to that described previously [29] . The T . cruzi Dm28c clone , classified in the TcI phylogenetic group [30] , was maintained in Novy-MacNeal-Nicolle media ( NNN ) and brain heart infusion media ( BHI- DIFCO ) supplemented with bovine serum albumin ( BSA ) and hemin . For the in vivo and ex vivo experiments , epimastigotes were collected during the exponential growth phase , washed three times in 0 . 15 M NaCl , 0 . 01 M phosphate-buffer , pH 7 . 2 ( PBS ) and used immediately [11] , [31] . R . prolixus were fed and raised according to the Ethical Principles in Animal Experimentation approved by the Ethics Committee in Animal Experimentation ( CEUA/FIOCRUZ ) under the approved protocol number P-54/10-4/LW12/11 . The experiments performed with citrated human blood using an artificial apparatus were conducted according to the Ethical Principles in Animal Experimentation approved by the Ethics Committee in Animal Experimentation ( CEUA/FIOCRUZ ) under the approved protocol number L-0061/08 . All blood donors provided informed written consent . Both protocols are from CONCEA/MCT ( http://www . cobea . org . br/ ) , which is associated with the American Association for Animal Science ( AAAS ) , the Federation of European Laboratory Animal Science Associations ( FELASA ) , the International Council for Animal Science ( ICLAS ) and the Association for Assessment and Accreditation of Laboratory Animal Care International ( AAALAC ) . Epimastigotes ( 109 ) were delipidated using a water/chloroform/butan-1-ol treatment and further extracted with butan-1-ol at 4°C as described previously [32] . Briefly , the soluble fraction was evaporated under an N2 stream , and the insoluble material was re-extracted with 66% butan-1-ol in water . The butan-1-ol phase ( F1 ) contained mainly lipids , phospholipids and glycoinositolphosphates ( GIPLs ) , whereas the aqueous phase ( F2 ) is enriched in mucins [32] . Both phases were further extracted with 9% butan-1-ol in water . Delipidated parasite pellets were also extracted with 9% butan-1-ol in water and the mucin-rich aqueous ( F3 ) and butan-1-ol ( F4 ) phases were stored . The final parasite pellets were resuspended in denaturing loading buffer containing 6 M urea and 100 µg/ml DNAse I ( SIGMA ) . In order to enrich in glycoconjugates , pellets containing 108 parasites were homogenized in ConA buffer ( 50 mM Tris-HCl , pH 7 . 4 , 150 mM NaCl , 1% NP40 , 0 . 1% Na deoxycholate , 1 mM PMSF , 50 µM TLCK , 1 mM DTT ) and fractionated in batch using 200 µl of ConA-sepharose ( GE Healthcare ) [26] . Elution was carried out with 300 µl of ConA buffer with 0 . 5 M α methylmannoside ( Sigma , St . Louis , MO ) . Parasite total lysates were treated with PI-PLC and submitted to Triton X-114 partition as described [26] , to ascertain the presence of GPI anchor . Gel electrophoresis was performed under denaturing conditions in 15% SDS-PAGE . For Western blots using total proteins , lysates corresponding to ∼107 parasites prepared as described [26] were loaded in each lane , transferred to PVDF membranes ( GE Healthcare ) , reacted with the appropriate antiserum followed by HRP-conjugated secondary Abs ( Sigma ) and developed using chemiluminescence ( Pierce ) . Antibodies to TcSMUG L were affinity-purified and used as described by [26] . Rabbit antiserum to glutamate dehydrogenase from T . cruzi ( TcGDH ) was used at 1∶3 , 000 dilution [33] . Peptides used in this study were synthesized bearing an acetyl group on their N-termini and a C-terminal Cys residue ( GenScript ) . Sequences were derived from the predicted N-terminal region of mature TcSMUG L ( AVFKAAGGDPKKNTTC ) , TcSMUG S ( VEAGEGQDQTC ) and TSSA ( TPPSGTENKPATGEAPSQPGAC ) products . When indicated , peptides were synthesized with a biotin group instead of the acetyl group on their N-termini . Although bioinformatics methods indicate that the sequences EEGQYDAAVFAVFKAAGGDPKKNTT and EEGQYDAAVFVEAGEGQDQT constitute the predicted mature N-termini for TcSMUG L and S products , respectively [26] , mass spectrometry-based data using purified epimastigote total mucins [34] , strongly suggested a further trimming of the EEGQYDAAVF sequence in vivo . After washing in PBS , epimastigotes were suspended in fresh BHI to a density of 2 . 5×107 cells/ml . Samples of an interaction medium composed of 200 µl of this parasite suspension together with posterior midguts , freshly dissected and washed only in PBS , from insects collected 10 days after a non-infectious blood meal , were placed in Eppendorf microtubes [10] and incubated for 30 min at 25°C ( non-treated control group ) . Under these conditions , epimastigotes adhered to the luminal surface of midgut epithelium cells [11] . For the experimental groups , the midguts were previously incubated ( 30 min , 25°C ) in PBS supplemented with TcSMUG S ( negative control ) , TcSMUG L or TSSA peptides at different concentrations . The treated-posterior midguts were then washed in fresh PBS and immediately added to the BHI interaction medium containing parasites . After incubation ( 30 min , 25°C ) , all midgut preparations were spread onto glass slides to count the number of attached parasites . A Zeiss microscope with reticulated ocular , equipped with a video microscopy camera , was used for counting parasites attached to 100 randomly chosen epithelial cells in 10 different fields of each midgut preparation . For each experimental group , 10 insect midguts were used [35] , [36] . Fifth-instar nymphs of regularly fed R . prolixus , which had been starved for 7 days after the last ecdysis , were fed on artificial bloodmeal apparatus with a mixture of heat-inactivated citrated human blood and epimastigotes ( 2×105 parasites/ml ) as previously described [37] . TcSMUG S ( negative control ) , TSSA or TcSMUG L peptide was added to the infected blood meal to a final concentration of 30 µg/ml just before feeding . At days 7 , 14 or 21 , the entire digestive tracts consisting of anterior midgut ( stomach ) , posterior midgut and rectum of 10 insects were dissected and homogenized in a small volume of PBS . Afterwards , additional PBS was added to fill the homogenates to 1 ml [38] , [39] . The number of parasites in each homogenate was determined using a Neubauer hemocytometer [40] , [41] . Each experiment was repeated at least three times . Posterior midgut compartments obtained by dissection were fixed for 2 h at room temperature in 2 . 5% glutaraldehyde diluted in 0 . 1 M cacodylate buffer , pH 7 . 2 , and washed twice in the same buffer . Post-fixation was performed in the dark for 2 h in 1% osmium tetroxide diluted in 0 . 1 M cacodylate buffer , pH 7 . 2 , followed by dehydration with continuous acetone series ( 70% , 90% and 100% , respectively ) . Samples were then embedded in epoxy resin and polymerized at 60°C for three days . Thick plastic sections were stained with toluidine blue and observed under an Axioplan MC 100 spot microscope [10] . Dissected posterior midgut fragments were fixed for 1 h at room temperature in 4% p-formaldehyde diluted in 0 . 1 M cacodylate buffer , pH 7 . 2 . Afterwards , samples were washed in PBS containing 1% of BSA , pH 7 . 2 ( PBS-BSA ) and incubated for 30 min in 50 mM ammonium chloride solution followed by another washing step in PBS-BSA at room temperature . Tissues were then incubated with biotin-labeled TcSMUG S , TSSA or TcSMUG L peptide diluted in PBS-BSA for 1 h at room temperature and washed again in PBS-BSA before incubation with FITC-labeled-Avidin conjugate ( SIGMA ) ( 1∶100 ) for 1 h and washed in distilled water in the dark for 10 min [42] . For the control groups , the incubation with biotin-labeled peptides was omitted . Finally , the tissues were spread onto glass slides for visualization using an emission filter of 488 nm and observed under an Axioplan MC 100 spot microscope coupled to an Axiovision system computer [43] . Results were analyzed using ANOVA and Tukey's tests [44] using Stats Direct Statistical Software , version 2 . 2 . 7 ( StatsDirect Ltd . , Sale , Cheshire , UK ) . Differences between treated- and control-groups were considered non-statistically significant when p>0 . 05 . Probability values are specified in the text .
Previous results indicate that the expression level of TcSMUG L-encoded products is quite variable among epimastigotes from different T . cruzi isolates [26] . Therefore , as a first step toward the validation of our R . prolixus infection model , we undertook preliminary characterization of TcSMUG L products in the DM28c stock . Western blotting assays carried out over total epimastigote lysates and probed with affinity-purified antibodies directed against an N-terminus-derived TcSMUG L peptide revealed a major ∼35 kDa band , thus in the range of fully processed TcSMUG L products described in other parasite stocks [26] ( Fig . 1A ) . As controls , we used analogous fractions from epimastigotes from Adriana and CL Brener stocks , which showed the greatest differences in terms of TcSMUG L expression [26] . The results were normalized by re-probing the membrane with antiserum directed against TcGDH . Densitometric analyses indicated that TcSMUG L expression levels from the DM28c stock were roughly equivalent ( 86% ) to that of CL Brener . These products were removed from the parasite surface following PI-PLC treatment [26] , a molecular signature of GPI-anchored molecules ( not shown ) , and were specifically retained following ConA chromatography ( Fig . 1B ) , indicating they bear terminal α-D-mannosyl and/or α-D-glucosyl residues , as described for other stocks [26] . To analyze whether TcSMUG L products behaved as mucin-type proteins , i . e . , underwent extensive O-glycosylation , we purified total mucins from Dm28c epimastigotes following a standard butan-1-ol extraction protocol [32] and probed these fractions by Western blot . As shown in Fig . 1C , TcSMUG L products were mostly detected in the F3 fraction , which was highly enriched in gp35/50 , as verified by mAb 2B10 and 10D8 reactivity ( not shown ) . The presence of high-molecular weight aggregates in purified TcSMUG L products has been described for other T . cruzi mucin-type glycoconjugates [22] , [26] . A minor fraction was also revealed in the pellet , which might be ascribed to incomplete extraction . Together , these results strongly suggest that Dm28c epimastigotes express high levels of fully processed TcSMUG L product on their surface . To assess whether TcSMUG L products can act as direct ligands for possible receptors in insect epithelial midgut cells , we tested the effect of pre-treatment of dissected midguts with a peptide spanning the TcSMUG L mature N-terminus . As controls , we assayed in parallel the effect of the corresponding peptide derived from TcSMUG S and TSSA , a member of the TcMUC family of mucins . As a first set of experiments , in posterior R . prolixus midgut preparations obtained from a control ( non-treated ) group , 114 . 8±28 . 2 epimastigotes were found attached per 100 midgut cells ( Fig . 2A ) . Similar adhesion rates ( 128 . 8±34 . 7/100 midgut cells ) were obtained when midguts were first incubated with 1 µg/ml of a control TcSMUG S peptide ( p>0·05 ) ( Fig . 2A ) . In contrast , attachment of only 28 . 5±28 . 4 and 20 . 8±10 . 06 epimastigotes per 100 cells of the midgut epithelium were recorded when the flagellates were pre-incubated with 1 µg/ml of either TcSMUG L or a control TcMUC-derived ( TSSA ) peptide ( p<0·0001 ) , respectively ( Fig . 2A ) . A dose-dependent effect on the ex vivo attachment of epimastigotes was verified for the latter molecules , indicating that the presence of either synthetic peptide blocked a potential ligand-receptor interaction involved in epimastigote attachment ( Fig . 2B ) . As shown in Fig . 2B , incubation with 0 . 01 µg/ml of the TcSMUG L peptide did not affect flagellate adhesion rates when compared with the control group , whereas incubation with 0 . 1 µg/ml or 1 µg/ml of the TcSMUG L peptide reduced T . cruzi attachment to 40 . 8 ±16 . 78 and 30 . 8 ±10 . 42 ( p<0·01 ) epimastigotes per 100 midgut cells , respectively . Similarly , midgut incubation with 0 . 01 µg/ml of the TSSA peptide resulted in 128 . 6±20 . 87epimastigotes attached per 100 midgut cells and did not affect flagellate adhesion rates when compared with the control group ( 123 . 2±23 . 74 epimastigotes/100 midgut cells ) , whereas incubation with 0 . 1 µg/ml or 1 µg/ml of the same peptide reduced T . cruzi attachment to 37 . 6 ±19 . 65 and 30 . 6 ±12 . 4 ( p<0·001 ) epimastigotes per 100 midgut cells , respectively ( Fig . 2C ) . Therefore , our results showed that the pre-incubation of R . prolixus midguts with the TcSMUG L or TSSA peptide promote significant alteration of the epimastigote-midgut interaction rate . Upon ingestion of approximately 2×105 Dm28c epimastigotes/ml of blood , fifth-instar nymphs of R . prolixus became heavily infected with T . cruzi ( Fig . 3 ) . In the control group , the infection levels varied from 3 . 33±0 . 35×105 flagellates/ml of digestive tract homogenate 7 days after infection to 2 . 06±0 . 10×106 flagellates/ml of digestive tract 21 days post-infection . Similar infection levels were observed throughout the time frame of the experiment in insect groups fed with blood supplemented with either TcSMUG S or TSSA peptide ( p>0·05 ) . In contrast , nymphs fed with blood supplemented with TcSMUG L peptide showed significantly reduced infection levels . Direct counts revealed 2 . 3±0 . 12×102 ( p<0·0001 ) and 2 . 3±0 . 27×102 ( p<0·0001 ) flagellates/ml of digestive tract homogenate 14 and 21 days post-infection , respectively , representing a ∼4-log difference from controls . Even more compelling , no parasites were observed 7 days post-infection in TcSMUG L peptide-treated insects . Together , these results suggest that soluble TcSMUG L peptide significantly inhibits the normal development of Dm28c parasites in R . prolixus , likely by interfering between the interaction of endogenous TcSMUG L products displayed on the surface of epimastigotes and triatomid midgut receptors . Light microscopy of R . prolixus midgut showed a single columnar epithelium composed by posterior midgut cells . Toluidine-stained granules were observed in the apical and medial region , where a round nucleus was located . As previously described [10] , these epithelial cells were closely joined at their medial and basal regions , whereas a brush border associated with the PMM was observed at the luminal surface of their apical regions ( Fig . S1 ) . No significant labeling was obtained after incubation of R . prolixus posterior midgut surface with Avidin-FITC conjugate alone ( Fig . 4A , B ) or after previous incubation with biotin-labeled TcSMUG S peptide followed by the Avidin-FITC conjugate ( Fig . 4E , F ) . However , in line with previous results , fluorescence of specific binding sites was observed on the surface of luminal posterior midgut cells after pre-incubation with biotin-labeled TcSMUG L ( Fig . 4C , D ) or TSSA ( Fig . 4G , H ) peptide under the same conditions . Unexpectedly , the samples pre-incubated with TSSA also showed some intracellular staining , particularly in the nucleolus , which may be attributed to partial permeabilization of the cells during fixation .
During its life cycle , T . cruzi adheres to specific host molecules/cell types as essential steps for parasite survival . Depending on the parasite developmental stage and the nature of the involved molecules , these interactions trigger a variety of events such as bidirectional cell signaling , host cell internalization , parasite replication or transformation to infective stages [45] , [46] . Within the triatomid vector , different lines of research have established that molecules able to inhibit parasite attachment to insect tissues ex vivo also often efficiently block the in vivo development of T . cruzi [35] . For instance , purified GIPLs were shown to bind to the luminal surface of the posterior midgut . Accordingly , their exogenous addition dramatically impaired both ex vivo attachment of epimastigotes to this organ and the flagellate multiplication in the insect digestive tract , which prevented the successful colonization of the vector [11] . Similar effects were described for different carbohydrate-binding proteins ( CBPs ) of the epimastigote surface with a strong affinity for higher glycan oligomers and sulfated glycosaminoglycans ( S-GAGs ) present in the posterior midgut of R . prolixus [36] , [47] , [48] . The net negative charge of both S-GAGs and specific carbohydrates may act as a first , non-specific step prior to T . cruzi adhesion to specific receptors in the luminal midgut PMM [35] . In addition , an antiserum raised against R . prolixus PMM and midgut tissue interfered with midgut structural organization and slowed the development of T . cruzi in the insect vector [49] . The entire surface , including the cell body and the flagellum , of various T . cruzi developmental forms is covered with mucins that play a key role in parasite protection [50]–[52] , infectivity , and development [15] . T . cruzi mucins are anchored to the outer leaflet of the plasma membrane through a GPI motif and undergo extensive glycosylation in their central Thr-rich domain . These features confer strong hydrophilic characteristics and an extended ( “rod-like” ) structural conformation [53] , which is often used to elevate an outermost peptide above the parasite glycocalix . This N-terminal peptide , which is not predicted to be O-glycosylated , is thus ideally suited to participate in cell-to-cell interaction phenomena [54] . The results presented here strongly suggest that the N-terminal peptide of TcSMUG L products is required for efficient interaction between the parasite and the insect midgut and the subsequent growth of the flagellate in the invertebrate host . As shown , addition of the exogenous peptide led to a significant reduction in ex vivo adhesion to the insect midgut , and also inhibition of in vivo development within vectors . Due to its small molecular size , this effect is unlikely to be caused by steric effects , where the TcSMUG L peptide would prevent access of parasite recognition molecules to specific sites in the insect gut cells . Quite the opposite , we favor the hypothesis that the exogenous TcSMUG L peptide exerts its inhibitory effect by outcompeting the parasite binding sites in the triatomine luminal surface of the midgut epithelium . This idea is further supported by histochemical data showing intense labeling of the surface of luminal posterior midgut cells after pre-incubation with biotin-labeled TcSMUG L peptide . Therefore , it is likely that TcSMUG L products act as surface adhesion molecules , promoting epimastigote adhesion and colonization through recognition of specific receptor ( s ) on insect cells . In this framework , a distinct expression profile verified for TcSMUG L products [26] could contribute to the biological heterogeneity found between different isolates of T . cruzi in terms of triatomid infectivity . Moreover , drastic reduction in TcSMUG L expression upon differentiation to metacyclic trypomastigotes suggests a developmental regulation program that could help to explain why these latter forms are detached from the midgut surface [26] . One unexpected and puzzling finding was that the exogenous TSSA-derived peptide showed adhesion properties to insect midgut cells , as well as ex vivo inhibition on epimastigote attachment . It is worth mentioning that TSSA belongs to the TcMUC group of genes , which is expressed during the mammalian-dwelling stages of the protozoan [20] , [21] , [54] . In particular , TSSA expression is restricted to the surface of blood trypomastigotes , the parasite stage ingested by the vector during an infective blood meal , and amastigote-to-bloodstream trypomastigote intermediate forms . From a structural staindpoint , and despite showing similar bias in amino acid composition ( with Cys , Phe , Trp and Tyr amino acids -all residues that could perturb the physicochemical properties of T . cruzi mucins- being underrepresented or absent ) , there are no obvious similarities in the primary sequences of the TSSA and TcSMUG L peptides that could explain their similar binding properties . Indeed , the labeling pattern obtained for TSSA in posterior midgut sections is different than that obtained for the TcSMUG L peptide , suggesting they recognize different receptor ( s ) on the surface of insect cells , although more studies would be required to address this point . Importantly , and in strict correlation with its expression profiling , the interaction between TSSA and insect midgut cells seems to have no biological relevance , as it had no effect on parasite in vivo development . Although little is known about the mechanisms leading to the remodeling of the surface coat when the flagellate moves from the mammal into the insect vector , it is reasonable to suppose that TSSA is shed during this process . Free in the insect stomach , TSSA may reach the posterior midgut and be recognized by PMM receptors for mucins or other glycoconjugates . Transfer of antigenic epitopes from T . cruzi to the PMM of Triatoma infestans has been previously described [55] . In spite of this , TSSA does not seem to participate in the protozoan development of R . prolixus , which is compatible with its lack of expression in insect-dwelling stages of T . cruzi . Altogether , these findings establish that TcSMUG L products are involved in the interaction between T . cruzi and its invertebrate host . Indeed , our results demonstrate that these products are involved in successful adhesion to the epithelial cells of insect vectors both ex vivo and in vivo , although the exact molecular mechanism , and particularly the putative receptor on the surface of the insect cells , should be further explored . Most importantly , a severe reduction in flagellate population in the digestive tract of R . prolixus was observed when triatomines were infected with epimastigotes of T . cruzi and simultaneously orally treated with the TcSMUG L peptide . Collectively , our work adds new insight into the relevance of mucin-type glycoconjugates in the infection of insect vectors and points to them as promising targets to develop transmission-blocking strategies for this disease . | Chagas disease , the major tropical human disease in much of Latin America , affects approximately 11 million people . There are 300 , 000 new cases of Chagas disease and approximately 21 , 000 deaths , annually . Triatomine vectors , including Rhodnius prolixus , are able to transmit the protozoan Trypanosoma cruzi , the etiological agent of disease . To develop within insects , the flagellates undergo morphological changes , modulating surface molecules to enable interactions with insect tissues such as the perimicrovilar membranes in the midgut which is an essential step for their development and successful transmission to a vertebrate host . The surface of T . cruzi is covered in glycosyl phosphatidylinositol ( GPI ) -anchored mucin molecules that determine parasite protection and establishment of a persistent infection in vertebrates . A particular kind of mucin , termed TcSMUG L , is only present at surface of the insect-dwelling stages of protozoan and , according to our results , it is involved in the interaction between T . cruzi and its invertebrate host , determining both the ex vivo adhesion to the insect midgut cells and the in vivo development in the vector . Collectively , our work adds new insight into the relevance of mucin-type glycoconjugates in the infection of insect vectors and points to them as promising targets to develop transmission-blocking strategies for this disease . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [] | 2013 | Trypanosoma cruzi TcSMUG L-surface Mucins Promote Development and Infectivity in the Triatomine Vector Rhodnius prolixus |
The growing burden of dengue in many countries worldwide and the difficulty of preventing outbreaks have increased the urgency to identify alternative public health management strategies and effective approaches to control and prevent dengue outbreaks . The objectives of this study were to understand the impact of dengue outbreak on different stakeholders in Brazil , to explore their perceptions of approaches used by governmental authorities to control and prevent dengue outbreaks and to define the challenges and implications of preventing future outbreaks . In 2015 , a qualitative study was conducted in two urban states in Brazil: São Paulo , which was experiencing an outbreak in 2015 , and Rio de Janeiro , which experienced outbreaks in 2011 and 2012 . Face-to-face interviews using a semi-structured questionnaire were conducted with nine different categories of stakeholders: health workers ( physicians , nurses ) , hospital administrators , municipal government representatives , community members and leaders , school administrators , business leaders and vector control managers . Interviews were focused on the following areas: impact of the dengue outbreak , perceptions of control measures implemented by governmental authorities during outbreaks and challenges in preventing future dengue outbreaks . A total of 40 stakeholders were included in the study . Health workers and community members reported longer waiting times at hospitals due to the increased number of patients receiving care for dengue-related symptoms . Health workers and hospital administrators reported that there were no major interruptions in access to care . Overall financial impact of dengue outbreaks on households was greatest for low-income families . Despite prevention and control campaigns implemented between outbreak periods , various stakeholders reported that dengue prevention and control efforts performed by municipal authorities remained insufficient , suggesting that efforts should be reinforced and better coordinated by governmental authorities , particularly during outbreak periods . The study shows that a dengue outbreak has a multisectorial impact in the medical , societal , economic and political sectors . The study provides useful insights and knowledge in different stakeholder populations that could guide local authorities and government officials in planning , designing and initiating public health programs . Research focused on a better understanding of how communities and political authorities respond to dengue outbreaks is a necessary component for designing and implementing plans to decrease the incidence and impact of dengue outbreaks in Brazil .
Since the beginning of the 21st century , dengue fever has been a significant vector-borne arboviral disease , occurring mainly in tropical and sub-tropical countries where more than 3 . 9 billion people are at risk of infection in 128 countries [1–4] . With an estimated 400 million annual dengue incidence worldwide , the disease is currently endemic in more than 125 countries in Africa , the Americas , South East Asia and other regions in the world [2] . Sixty percent of dengue cases occur in the Americas , predominantly Latin America , where the disease has re-emerged owing to re-infestation by the dengue vector . Since its re-emergence in Latin America , dengue has spread dramatically throughout the region [5] . The number of reported dengue cases rose from 1 , 033 , 417 in the 1980s to 2 , 725 , 405 in the 1990s , and 4 , 759 , 007 between 2000 and 2007 [2 , 5] . Between 2001 and 2009 , six countries accounted for more than 75 percent of all cases in the region: Venezuela , Brazil , Costa Rica , Colombia , Honduras and Mexico [6 , 7] . Various political , environmental and social factors influenced the re-emergence of the disease in Latin America . Dengue is largely an urban disease [7] . Overcrowding , uncovered water sources and climate change are optimal conditions for geographic spread of the mosquito vector for dengue , which lives in domestic settings and can breed in very little water and survive drought conditions [1] . The impact of urbanization is particularly significant in low and middle-income countries [8–10] . Among the South American countries of Argentina , Brazil , Chile , Paraguay and Uruguay , Brazil has the highest incidence rate of dengue with 294 cases per 100 , 000 inhabitants in 2014 [11] . Brazil is considered a tropical country in its entirety because of its hot and humid climate , which provides a favorable environment for proliferation of the dengue vector . Incidence of dengue in Brazil has frequently been high , and the number of cases in the country has sometimes represented as much as 60% of all reported dengue cases worldwide [11] . After the disruption of its vector control program , Brazil experienced a series of dengue outbreaks [5 , 11] characterized by increasing geographic spread and severity as estimated by the number of hospitalizations and deaths [11] . In the first eight months of 2015 , Brazil experienced an increase in dengue of 140 percent over the same period during the previous year ( 1 , 416 , 179 cases from January to August 2015 versus 589 , 107 cases in 2014 ) . During that same period , the number of severe cases increased dramatically ( 1 , 284 severe cases in 2015 versus 664 severe cases in 2014 ) as did deaths ( 693 deaths in 2015 versus 407 deaths in 2014 ) [11] . Dengue imposes significant economic and societal burdens on countries where the disease is endemic and , as such , estimating the associated disease impact can help inform policymakers and assist them in setting priorities for disease control and management strategies [12 , 13] . The effects of dengue on health and preventive care , its economic burden and social impact on populations have not been clearly studied . Understanding dengue burden from societal and socio-economic perspectives is crucial for allocation of limited scarce public health resources among competing health threats , as well as ensuring cost-effectiveness of integrated dengue prevention and control methods . One such method promoted by the World Health Organization to overcome challenges associated with conventional single-intervention approaches is Integrated Vector Management ( IVM ) . Defined as a “rational decision-making process for the optimal use of resources for vector control , ” IVM considers five key elements in the implementation of IVM to prevent vector-borne diseases such as dengue: advocacy , social mobilization and regulatory control; collaboration within the health sector and with other sectors; integration of non-chemical and chemical vector control methods; evidence-based decision-making; and development of adequate human resources , training and career structures [14] . The growing burden of dengue and the challenges of preventing dengue outbreaks ( including increase in prevention costs such as on vector control ) have increased the urgency of the public health sector to identify alternative management strategies [15–19] . Monitoring of public impact , perceptions and behaviors is needed in order to guide the development of adequate and effective strategies to controlling and preventing dengue outbreaks . Establishing key performance indicators is important in measuring the effectiveness of existing and current surveillance and vector programs . In addition , identifying and understanding societal , cultural and environmental factors related to dengue outbreaks may provide insight for the development of targeted dengue prevention and control interventions , including community involvement and engagement in vector control activities given that community-based interventions have been found to impact vector indices [20] . Furthermore , understanding dengue vector-virus-disease behaviour and human-animal-environment interactions and interdependencies , known as “One Health , ” is critical to IVM [21] . This can provide early and timely added value for consolidated and harmonized mitigation and adaptation tactics and advocacy in local settings [22] . Given the dearth of literature , qualitative studies are needed to gain in-depth understanding of stakeholder and community reactions to a dengue outbreak and its related impact . The growing burden of dengue in Brazil presents an interesting model for qualitative research around the challenges and issues that need to be addressed to strengthen dengue prevention and control interventions in Brazil and throughout Latin America . The objectives of this study were to understand the impact of dengue outbreak on different stakeholders in Brazil , to explore stakeholder perceptions of approaches used by governmental authorities to control and prevent dengue outbreaks and to define the challenges and implications of preventing future outbreaks .
This qualitative study was conducted from April to August 2015 . The study sites for collecting responses to interviews with stakeholders were the municipalities of Marília and Sorocaba in São Paulo State and the municipalities of Búzios and Rio de Janeiro in Rio de Janeiro State . Both states are densely populated , urban areas located in southeast Brazil , characteristics favorable for dengue outbreaks . Both of these states experienced recent epidemic dengue outbreaks , defined as at least 300 cases per 100 , 000 inhabitants [19] . The 2015 outbreak in São Paulo State was concurrent with the study and had an incidence of 3 , 762 and 9 , 072 inhabitants per 100 , 000 in the municipalities of Marília and Sorocaba , respectively . Rio de Janeiro State experienced a series of outbreaks that had surmounted by the time of the study . Rio de Janeiro experienced outbreaks with incidences of 1 , 090 . 5 per 100 , 000 inhabitants in 2011 and 567 . 7 per 100 , 000 inhabitants in 2012 [11] . Búzios experienced outbreaks in 2007 and 2011 with incidences of 2 , 007 and 1 , 994 inhabitants per 100 , 000 , respectively , for each year [11] . A convenience sampling method was used to select target participants at the two research sites . With the goal to obtain responses from a diverse set of representative stakeholders affected by dengue outbreaks , various stakeholders were defined and recruited . A sample of 30–50 stakeholders across the two sites was determined to be sufficient to qualitatively address the study objectives . Stakeholder categories were selected based on the type of individuals expected to best address questions related to the study objectives . Nine categories of stakeholders were identified: municipal government representatives , hospital administrators , physicians , nurses , community members and leaders , school administrators , business leaders and vector control managers . The categories of hospital administrators and health workers ( physicians and nurses ) were subdivided into two sectors , public and private . Stakeholders were identified from communities in each study site particularly affected by a dengue outbreak based on incidence data reported in local Ministry of Health dengue bulletins and community recommendations . Specific stakeholders in public and private hospitals within those communities were targeted based on their level of involvement in the dengue outbreak response . Community members and leaders , school administrators and business leaders were targeted to capture a diversity of opinion , particularly as it pertains to economic background . A semi-structured questionnaire was used for interviews , which focused on a range of topics regarding the impact of dengue outbreaks and response to the outbreak . Data were collected through face-to-face interviews conducted by three Brazilian social researchers . These social researchers were trained by Axios International and were based in each study site . Interviews focused on impact of dengue outbreaks and stakeholder opinion of dengue outbreak prevention and control efforts implemented by governmental authorities . The impact of dengue outbreaks was assessed in the following areas: healthcare infrastructure and case management , municipal government operations and finances , household operations and finances , and communities ( including schools and businesses ) . The duration of the interviews ranged from 30 to 40 minutes . All the interviews were audio recorded , then transcribed and translated from Portuguese to English . Interviews were conducted in stakeholder settings and at places that were convenient for the participants , such as their home or workplace . The data on costs of dengue prevention were obtained from Brazilian Government websites [23 , 24] . A directed content analysis approach was used to analyze the data and to identify the key themes . Audio transcriptions and interview notes were reviewed and summarized by two members of the study team . They reviewed all qualitative comments extracted from the questionnaires and placed them into broader categories based on content and theme . All data were systematically analyzed . Key findings and quotes were compiled in a Microsoft Excel table and coded according to topic or theme . Responses were reported if at least two persons gave a similar response . Final decisions on comment categories were discussed with a third author . The research team identified quotes that best illustrated common themes and included these quotes in the results of this study . A certified translator translated quotes into English . Because the results of interviews were quite similar across both states ( and the four municipalities ) , the decision was made to pool the results . Participation in this study was voluntary and all participants provided written informed consent prior to the interviews . All information was collected anonymously and the outcomes were used only for research purposes . Data of the interviews are housed in Axios International , Paris , France . The Western Institutional Review Board ( WIRB ) approved the research ( #1-904491-1 ) .
A total of 40 stakeholders were included in the study , with 18 stakeholders in São Paulo State ( Marília and Sorocaba municipalities ) and 22 in the Rio de Janeiro State ( Búzios and Rio de Janeiro municipalities ) . The different stakeholders included in the study were physicians ( Phy ) , nurses ( Nu ) , hospital administrators ( HA ) , community members ( CM ) , community leaders ( CL ) , school administrators ( SA ) , municipal policy makers ( MPM ) , vector control managers ( VCM ) and business leaders ( BL ) ( Table 1 ) . The study included three physicians and three nurses from a public facility and one physician and one nurse from a private facility . Two hospital administrators were from a private facility and one from a public facility . One nurse and one physician from separate private hospitals also had previous experience working in a public emergency care center during a dengue outbreak .
The study shows that a dengue outbreak undermines medical , social , societal , economic and political stability through a rapid patient influx into the healthcare system , the financial burden on governments and households , and the effect of community dissatisfaction and disengagement on outbreak control measures . This multi-sectorial impact , although limited , results in social and political disruption . In addition to providing a description of the challenges that need to be surmounted during dengue outbreaks , this study suggests important considerations that should be accounted for in the design and implementation of future dengue prevention and control interventions . For the majority of stakeholders interviewed in this study , dengue control and prevention efforts implemented by the authorities were seen as insufficient . As citizens witnessed the rapid growth of dengue cases in their communities , fear and anxiety for contracting the disease impacted social structures . Stakeholders expressed frustration over the lack of concerted and timely government efforts to prevent and control outbreaks in the community , and many did not perceive the impact of such interventions . Stakeholders also felt that prevention efforts did not appear to be sufficiently coordinated . Insufficient governmental response or late responses to outbreak management appeared to generate a lack of confidence from the community . As a result , communities became increasingly disengaged and uninterested in vector control efforts in their household or neighbourhoods . Our findings underline the importance of permanent and strong social mobilization efforts and concerted community interventions in dengue prevention and control and the difficulty in maintaining mobilization efforts in the community over time [19] . In regards to healthcare system , the significant and rapid influx of patients placed a major burden on public and private health facilities as they struggled to find the space and the staff to handle demand and the means to minimize increasingly long patient waiting times . Planning around potential outbreak scenarios prior to an outbreak and improving dengue case management in health facilities during outbreak periods is a priority in ensuring improved clinical outcomes and quality of care . This is particularly critical in cities with limited infrastructure and human resources . The study provides a significant description of the challenges that need to be surmounted during future dengue outbreaks and suggests important multifocal approaches that should be developed to meet these challenges . Indeed , the results of this qualitative study mirror studies done in other regions in the world , such as Southeast Asia , Ecuador and Mexico , with regard to economic burden and effects on community and municipal functions [13 , 25 , 26] . Poor populations are greatly impacted and approaches to education need to be improved . A better community understanding of approaches to dengue control is important to their successful implementation [8 , 14 , 27 , 28] . However , development of these educational programs requires knowledge of the culture , organization and functioning of community members and local authorities [14 , 28 , 29] . Such knowledge can be obtained by exploring the community’s level of understanding of dengue outbreaks obtained from information communicated by governmental authorities , media ( newspapers , radio , TV ) and social networks . To better manage and increase confidence of community members in vector control actions , protocols specifically designed for the local environment that can be rapidly implemented at the beginning of an outbreak need to be further developed [30] . Planning should include improved approaches to educating community members and leaders about dengue , increasing the availability of dengue resources before outbreaks and maintaining necessary resources and services between outbreaks . During outbreaks , local sensitization campaigns must focus on community mobilization activities to eradicate mosquito breeding . Knowledge or awareness has been reported as being important for the success of dengue prevention and control efforts , and inadequate knowledge about dengue was found to be a major impediment to the successful implementation of approaches used to eliminate dengue . Importantly , studies indicate that dengue knowledge alone did not single-handedly translate to adoption of preventive measures . Instead , policy makers and planners must also focus sensitization campaigns on reducing barriers to behaviour change related to control of dengue fever among the population [16 , 31] , and encouraging communities to adopt preventive measures [16] . Such actions should increase community confidence in government responses to dengue outbreaks while decreasing fear and misconceptions . Increasing the confidence of community members would also be expected to decrease the chances of political breakdown , a concern of community and government leaders during an outbreak . Dengue prevention and control should include individuals , families and the wider community and encourage community participation to improve its chances for success . In a routine dengue prevention and control program , a community based environmental strategy showed a significant reduction in levels of Aedes infestation by 50–75% compared with a routine program without community involvement [19] . Comprehensive and enhanced dengue intervention strategies based on community engagement could have a significant and effective impact on dengue outbreak control , yet are highly dependent on the community’s perception of the severity of the disease . Intervention strategies should be built into existing health care systems and closely coordinated with the national dengue control program [29 , 30] . The cost of dengue control at the household level also needs to be addressed in relevant protocols and community mobilization activities , particularly with regard to the allocation of resources during an outbreak and with planning for poorer populations . In this study , lower income populations also reported more challenges in covering the cost of treatment , or difficulty paying indirect costs , such as transportation charges for hospital or health facility visits [13] . While this study was not intended to quantify the economic impact of the disease , the financial data gathered highlights the substantial investment needed to cover the cost of on-going prevention and control efforts . At least R$ 4 . 2 billion ( US$ 1 . 28 billion ) was spent on dengue control and prevention efforts between 2010 and 2014 in Brazil [23 , 24] . At the federal , state and municipal levels alike , the financial burden of dengue is significant and multi-sectoral . State and federal contributions may not always be sufficient to meet municipal prevention and control needs . For example , Sorocaba’s spending in 2015 for dengue control and prevention was an estimated R$ 11 million ( US$ 3 . 5 million ) for the full year , nearly twice the total amount distributed by the state to all 645 municipalities in São Paulo [32 , 33] . The results also emphasize the critical challenge around financial accountability for dengue control and prevention among levels of government and the limited impact of costly year-round campaigns . LR Carasco et al . demonstrated that the average economic impact of dengue illness in Singapore from 2000 to 2009 ranged between US$0 . 85 billion and US$1 . 15 billion , of which control costs constitute 42%–59% [34] . A 2013 study conducted in 12 countries in Southeast Asia showed that , there was an annual average of 2 . 9 million ( m ) dengue episodes and 5 , 906 deaths from 2001–2010 . The annual economic burden ( with 95% certainty levels ) was US$950 million ( US$ 610million—US$ 1 , 384million ) or about US$1 . 65 ( US$ 1 . 06 —US$ 2 . 41 ) per capita . A study conducted in four Brazilian regions estimated that the cost for dengue prevention and control for the epidemic season in 2012–2013 was substantial [35] . The annual national economic burden was US$ 164 million from the public payer perspective , but may be as high as US$ 447 million ( adjusted for underreporting ) . From the societal perspective , the economic burden was US$ 468 million , but may be as high as US$ 1 , 212 million [35] . In addition to the medical , economic , social and political impact of dengue , this study also found that another key factor contributing to the need to prioritize dengue is that it is a highly visible disease . This visibility is due to the fact that dengue often occurs in epidemics , which attracts media attention , stokes public fear and puts a strain on municipal budgets . This is in contrast to non-epidemic diseases , such as diarrheal disease and pneumonia , which attract less public or media attention [36] . For this reason , community engagement and satisfaction is particularly important in the dengue context . Furthermore , given the visibility of the dengue , the role of the media during dengue outbreaks should be better understood . Appropriate and effective management approaches , ready to deploy before a dengue outbreak occurs , would decrease the burden on health care facilities , increase the confidence of community members in government responses to outbreaks and would keep municipalities operating with increased efficiency . Increased confidence of community members would also be expected to improve community involvement and compliance with official approaches to dengue control . Improving community involvement and compliance during dengue outbreaks should decrease the intensity and duration of the outbreak and decrease the chances of subsequent outbreaks [37] . Recently , a long-term follow-up and integrated efficacy analysis of 35 , 000 children between the ages of 2 and 16 years in Asian Pacific and Latin American countries showed a reduction in dengue disease in the efficacy surveillance phase among children and adolescents who received a recombinant live attenuated tetravalent dengue vaccine . The vaccine was also associated with a lower risk of hospitalization and severe dengue overall up to 2 years after completion of the three-dose vaccination schedule among children 9 to 16 years of age [38] . Disease impact modelling has further indicated that if 20% of the populations in the 10 countries that participated in these efficacy studies were vaccinated , this could potentially reduce dengue burden by 50% in these countries [39] . Improved approaches to dengue management that include the considerations discussed in this study will continue to play a key role in a comprehensive prevention and control strategy . Future research should include gaining an understanding of how a vaccine would help the local community , including analysis of the cost- effectiveness of a vaccine program and the impact of vaccination as part of an integrated prevention strategy for dengue in endemic countries [35 , 40] . Initiation of a public dengue immunization program could serve as a long-term and visible commitment by leaders to protecting populations against the economic and human impact of this disease , and hence may contribute to regaining public trust in governments where public confidence has been affected by a dengue outbreak . This study presents limitations . First , the results should be considered in the context of the study sample size and the nine categories of stakeholders included , with an unequal distribution of stakeholders per category . The unequal distribution is partly associated with difficulties faced in identifying voluntary participants in select stakeholder categories directly involved in prevention and control measures given the sensitive nature of the topic at the time of the study . In certain stakeholder categories there may exist some limitations to data transferability , yet in most categories , data saturation was achieved . Secondly , since the study is qualitative and explorative , caution must be used with interpretation of the results . Further investigations will require other methodological approaches such as quantitative methods to confirm our findings . In addition , while the value of focus group discussions in presenting varying opinions and perspectives is recognized , the sample size of the study did not permit for such an approach . Thirdly , the sample of informants may not have included all major stakeholders in each municipality or have been representative of all stakeholders . As with all qualitative studies with open-ended responses , there is also the possibility of misunderstanding or biased interpretation of participant responses . Nevertheless , the strength of this qualitative study lies in the sample comprised of a large panel of stakeholders from a diverse socio-demographic background , which closely resemble the population usually involved in dengue control . The structure of the interviews , which allowed for probing and clarification of responses , and analysis were designed to minimize misinterpretation . In conclusion , the study provides useful insights from different stakeholder populations that could guide local authorities and government officials in planning , designing and implementing integrated public health programs and activities aimed at preventing and controlling dengue in Brazil . The consequences of a dengue outbreak reach far beyond the people infected , undermining medical , social , economic and political stability . Ultimately , an effective means of preventing dengue outbreaks that combines vector control and vaccination would reduce the social and political disruption and economic impact of the disease . The recent explosive outbreak of the Zika virus in South America reinforces the urgent need for more integrated research and public health strategies to control arboviral diseases [41 , 42] . Similar to the qualitative study reported in this paper , research that is designed to understand how a community and its local government responds to dengue outbreaks and two other arboviruses ( chikungunya , Zika ) circulating in Brazil , is a necessary component for designing and implementing plans to decrease the incidence of dengue outbreaks in Brazil . Such research will inform public policy by providing evidence-based recommendations to decrease the burdens of dengue outbreaks on both communities and governments | Since the beginning of the 21st century , dengue fever has been a significant vector-borne arboviral disease; actually more than 3 . 9 billion people are at risk of infection in 128 countries . Dengue has become an increasing public health concern in Latin America , especially in Brazil , which has the highest incidence rate of dengue . Researches are needed to gain in-depth understanding of stakeholder and community reactions to outbreak and to explore the societal impact of dengue outbreaks . In 2015 , a qualitative study was conducted in two urban states in Brazil , which experienced recent outbreaks . Longer waiting times at hospitals due to the increased number of patients receiving care for dengue-related symptoms were reported , but without interruptions in access to care . Various stakeholders reported that dengue prevention and control efforts performed by municipal authorities remained insufficient . The consequences of a dengue outbreak reach far beyond the patients , undermining medical , social , economic and political sectors . Research focused on a better understanding of how communities and political authorities respond to dengue outbreaks is a necessity for designing and implementing plans to control dengue outbreaks . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"medicine",
"and",
"health",
"sciences",
"physicians",
"medical",
"doctors",
"economic",
"analysis",
"municipal",
"government",
"tropical",
"diseases",
"geographical",
"locations",
"social",
"sciences",
"political",
"science",
"health",
"care",
"research",
"design",
"health",
"care",
"providers",
"neglected",
"tropical",
"diseases",
"governments",
"infectious",
"disease",
"control",
"research",
"and",
"analysis",
"methods",
"infectious",
"diseases",
"economic",
"impact",
"analysis",
"health",
"economics",
"south",
"america",
"dengue",
"fever",
"qualitative",
"studies",
"economics",
"brazil",
"people",
"and",
"places",
"professions",
"population",
"groupings",
"viral",
"diseases"
] | 2017 | Societal impact of dengue outbreaks: Stakeholder perceptions and related implications. A qualitative study in Brazil, 2015 |
Treatment with broadly neutralizing antibodies ( bNAbs ) has proven effective against HIV-1 infections in humanized mice , non-human primates , and humans . Due to the high mutation rate of HIV-1 , resistance testing of the patient’s viral strains to the bNAbs is still inevitable . So far , bNAb resistance can only be tested in expensive and time-consuming neutralization experiments . Here , we introduce well-performing computational models that predict the neutralization response of HIV-1 to bNAbs given only the envelope sequence of the virus . Using non-linear support vector machines based on a string kernel , the models learnt even the important binding sites of bNAbs with more complex epitopes , i . e . , the CD4 binding site targeting bNAbs , proving thereby the biological relevance of the models . To increase the interpretability of the models , we additionally provide a new kind of motif logo for each query sequence , visualizing those residues of the test sequence that influenced the prediction outcome the most . Moreover , we predicted the neutralization sensitivity of around 34 , 000 HIV-1 samples from different time points to a broad range of bNAbs , enabling the first analysis of HIV resistance to bNAbs on a global scale . The analysis showed for many of the bNAbs a trend towards antibody resistance over time , which had previously only been discovered for a small non-representative subset of the global HIV-1 population .
With around 36 . 7 million people living with HIV in 2015 and an incidence rate of around 2 . 1 million each year [1] , infections with HIV continue to be a major global health issue . However , despite more than three decades of research , there is neither a vaccine against nor a cure available for infection with HIV-1 . HIV-1 infected patients are usually treated with a highly active antiretroviral therapy ( ART ) . ART suppresses the replication of the active virus , but it is not capable of eliminating viral reservoirs and thus clearing the infection . To reduce the emergence of drug-resistant viruses , ART usually consists of a combination of three or more drugs from at least two different drug classes . In total , there are six different drug classes , which differ in their mode of interference with the HIV-1 life cycle , resulting in more than 20 available antiretroviral drugs . A change of the drug regimen is still often required , due to emerging drug resistances or side-effects . Since lifelong treatment is inevitable , for some patients no efficient drug regimens might be left eventually . Hence , there is still a high demand for drugs with new targets [2] . A currently investigated treatment option is the passive transfer of a combination of broadly neutralizing antibodies ( bNAbs ) to HIV-1 patients . Upon the advent of new single-cell antibody cloning techniques [3–5] and followed structure-based rational design approaches [6] , an abundance of these new bNAbs has been isolated and their higher neutralization potency and breadth have been shown in several studies [6–10] . The potency of an antibody is defined as the antibody concentration needed to inhibit HIV-1 infectivity by 50% ( IC50 ) or 80% ( IC80 ) , while the neutralization breadth of an antibody is measured by the ability of the antibody to neutralize viruses from different subtypes . The latter characteristic is very important in the case of HIV-1 due to its high molecular diversity within a patient but also within a population . The sole target of these neutralizing antibodies is the viral envelope glycoprotein , the so-called envelope spike , on the surface of the virus . The surface of the virus itself is made of host-lipids and is therefore undetectable by the immune system . Each spike consists of a trimeric heterodimer of two viral envelope glycoproteins , gp120 and gp41 , which are cleaved from the envelope glycoprotein , gp160 . While gp41 mediates host cell fusion , gp120 is essential for cell entry [11] . By successful binding of a neutralizing antibody to a spike , a chain reaction is initiated by the host immune system that eventually leads to the elimination of the virus . So far , there are five known sites on the envelope glycoprotein , which are targeted by a variety of bNAbs ( given in brackets ) : on gp120 the CD4 binding site ( e . g . , VRC01 , VRC-PG04 , 3BNC117 , NIH45-46 ) [9 , 12–14] , the V1/V2 region ( e . g . , PG9 and PG16 ) [7 , 8 , 15–17] , and the V3 loop ( e . g . , PGT128 , PGT121 , 10-996 , 10-1074 ) [8 , 10 , 18–21]; the membrane proximal external region ( MPER ) on gp41 ( e . g . , 10E8 ) [22–25]; and a newly identified site comprising parts of gp41 and gp120 ( e . g . , 35O22 ) [26] . Since the specific binding sites of bNAbs , so-called epitopes , on the envelope protein are not similarly accessed by any available drug , a therapy with bNAbs would offer a new effective treatment option for patients with resistance to all current therapies or might boost existing therapy combinations with few active drugs [27] . The efficacy of a treatment with a combination of these broad and potent neutralizing antibodies has been first shown in HIV-1 infected humanized mice [28 , 29] and non-human primates [30] . Tolerance and safety of the bNAbs VRC01 [31] and 3BNC117 [32] have been shown in phase 1 clinical trials in HIV-1 infected humans , where for 3BNC117 also the effective suppression of viremia could be observed . In addition , recent studies have shown that antiretroviral therapy with only one bNAb ( 3BNC117 ) is able to enhance the host immune response against HIV-1 [33] and leads to a significant delay of viral rebound after treatment interruption [34] . In contrast to ART , which usually requires a daily intake of the drugs , bNAbs have a longer half-life time , being able to control the viral load for more than 28 days in humans after administration [32] . High genetic variation of the viral envelope glycoproteins together with a glycan shielding of more conserved regions on the envelope often allow the virus to escape immune recognition [35] . Thus , for treatment success , neutralization resistances of the patient’s viral strains to the given bNAbs must be detected beforehand . Up to now , the neutralization sensitivity of a virus to an antibody can only be determined in time-consuming and expensive neutralization assays . To ensure a routine clinical practice , these tests have to be more rapid and cost-effective . This can be achieved , for example , by developing a genetic resistance test , coupled with a resistance prediction method similar to current decision support for ART treatment against HIV [36] . Since the envelope spike is the sole target of bNAbs , it is sufficient to consider the changes in the genetic composition of the viral envelope glycoproteins associated with changes in neutralization sensitivity of the virus . So far , the neutralization together with the genetic information has been mainly used to determine potential epitopes of bNAbs or to identify immunogens to elicit bNAbs . The aim of neutralization-based epitope prediction models is to learn potential epitopes or patches of the bNAb in the amino acid sequence of the envelope protein . There are approaches using only the neutralization information [37–41] or including structural information [41 , 42] . Changes in the amino acid composition of the epitopes are assumed to be associated with a change in neutralization sensitivity and thus can be learned from neutralization activity information . As a consequence , the model learns potential sites instead of predicting neutralization sensitivity . Nevertheless , some of the models , or more precisely the learnt sites , have been used to predict the neutralization activity for validation purpose . Unfortunately , the performance might be overoptimistic if the same data is used for learning the sites and the prediction task [42] . Another application is the identification of immunogens to elicit bNAbs . Therefore , Gnanakaran et al [43] compared the viral sequences of HIV-1 infected individuals with and without a broad and potent antibody response , hypothesizing that shared features among the viral sequences in individuals eliciting bNAbs might be potential immunogens . Shared features have been learnt using conditional mutual information together with an ensemble learning technique using classification trees . Similar to the above approaches , the identified features have been validated by predicting the neutralization sensitivity . An overview of a variety of computational approaches for epitope vaccine design is given by He et al . [44] . Recently , an artificial neural network approach has been proposed to directly model the IC50 value based on the envelope sequence information [45] . For this , the amino acids were mapped to integers . However , the authors modeled each position in the sequences as a continuous variable instead of a categorical one , which leads to a different interpretation of changes between different amino acids . In addition , only the performance of the older bNAb 2F5 was provided . IDEPI [46] is a very generic framework that , among other features , models the neutralization sensitivity of the virus to bNAbs using a linear support vector machine ( SVM ) and the envelope sequence of the virus . The above presented models have several shortcomings . First , potential epitopes can be poor immunogens . Second , sites outside the epitope can have an influence on the binding success of a bNAb as well , and thus also have an influence on the neutralization sensitivity . Structural information and other prior information about the binding sites might not be available for newly identified bNAbs . Most methods assume a linear relationship between changes in the amino acid composition and neutralization sensitivity on the one hand [37 , 46] and the independence of the epitope sites on the other [37] . This assumption might not hold for bNAbs targeting a more complex binding site . Another important point involves the handling of amino acid positions in the variable regions of the envelope protein . Though the variable regions are hard to align , they are also the regions where resistance mutations are likely to appear and thus these sites should not be dropped from the analysis [43] . In this study , we present prediction models for 11 different bNAbs ( VRC01 , VRC-PG04 , 3BNC117 , NIH45-46 , PG9 , PG16 , PGT121 , PGT128 , 10-996 , 10-1074 , and 35O22 ) that learnt discriminant signals ( amino acids or patterns of amino acids ) in the genetic sequence of the envelope glycoprotein gp160 ( envelope sequence ) , which influence the neutralization sensitivity to the particular antibody . To learn the neutralization susceptibility of HIV-1 strains to bNAbs , we trained our prediction models on data from three previously published neutralization assays [10 , 26 , 47] . Depending on the neutralization assay , IC50 titers for 115 to 220 HIV-1 isolates were available for each of the bNAbs . Following neutralization assay protocols , we used an IC50 value above 50 μg/mL as a threshold to determine neutralization resistance of a virus to a particular antibody . Based on the available IC50 titers for the HIV-1 isolates , the corresponding envelope sequences , and the threshold , we built binary classifiers with non-linear support vector machines ( SVM ) and string kernels to distinguish between HIV-1 resistance and susceptibility to a bNAb . As non-linear prediction models are often seen as black boxes , we trace back what each classifier learnt from the data and show that many of the learnt discriminant signals are known to play an important role for the binding success of the antibody . For a better interpretation of the classification decision ( resistant or susceptible ) , we provide a new way to produce motif logos that illustrate which and up to what extent amino acids in the tested sequence contributed to the particular classification result . Though we use the complete envelope sequence information , we show that only a few signals are important for the classification outcome and that models based only on these signals achieve comparable prediction power . To study the evolution of HIV-1 resistance to bNAbs , we additionally built regression models using support vector regression that directly predict the IC50 value from the envelope sequence of the virus . With these models we analyzed the neutralization sensitivity of HIV-1 to the considered 11 bNAbs for around 34 , 000 HIV-1 samples of different subtypes over a time period of more than 30 years from the Los Alamos HIV sequence database [48] . Thereby , we could not only confirm previous , experimental results , showing that there is a trend towards bNAb HIV-1 resistance over time in the subtype B population of HIV-1 on a much larger and more diverse data set , but for the first time , the trend could also be observed for the global HIV-1 population—a scale-up that would be very expensive in an experimental setting . A preliminary version of this study [49] has been published as a preprint .
In order to investigate whether neutralization sensitivity of HIV-1 to bNAbs has changed over time , we additionally built support vector regression models to directly predict the ( logarithmized ) IC50 value for the 11 considered bNAb . For subtype B variants , a continuous trend towards resistance has been already confirmed in certain cohorts ( around 40 samples ) of the French and Dutch HIV-1 population [55–57] . Since evolving resistance to antibody neutralization in the HIV-1 species would have major implications on the antibody selection for current vaccine development , it is important to know whether such a drift towards resistance also exists in the global HIV-1 population for all subtypes . In contrast to an experimental setting , where the large number of viral strains and the accruing costs make neutralization assays for the comprehensive global population hardly possible , our prediction models can be easily used to examine this question based on the vast amount of available sequence data . To model the global HIV-1 population over time , we used all available envelope sequences from the Los Alamos HIV sequence database ( around 34 , 000 after data processing , see Methods and S9 Table for accession numbers ) comprising viral isolates from all major subtypes over a time interval from 1981 to 2013 . We divided the given time interval into the following six time periods to account for changes in HIV-1 treatment strategies: 1981-1986 before ART , 1987-1991 ART monotherapy , 1992-1995 ART combination therapy ( cART ) , 1996-1999 cART with protease inhibitors , 2000-2005 cART with Lopinavir/Ritonavir , and 2006-2013 cART with Maraviroc/Raltegravir . With this partitioning of the data , we additionally covered the considered time intervals in the previously performed experimental studies [55–57] . An overview of the different subtypes and country distribution per time period are displayed in S2 Fig . In order to identify a drift towards resistance , we performed a permutation test for umbrella alternatives [58] on the predicted ( logarithmized ) IC50 values grouped by the six time periods . The umbrella test [59] is a more general test than the Jonckheere-Terpstra test [60 , 61] . Instead of testing for a monotonic trend , it tests for a peak in one of the time periods—a trend , monotonically increasing before and decreasing after the peak . The permutation test of umbrella alternatives [62] provides in additional partial p-values for each group , which enables a better analysis of the trend . Here , we define a trend towards resistance , if the peak is in the last time periods ( see Methods for details ) . In contrast to the experimental studies [55–57] , our data set is much larger , covers longer time periods , and is more heterogeneous . Thus , we expected to see more variation in our groups and therefore decided to use the umbrella test as a more general test in our case . However , we additionally provide the statistics for the Jonckheere-Terpstra test in S6 Table , which can be seen as a more conservative test . When considering only the subtype B variants of the around 34 , 000 viral isolates ( 17 , 392 ) , we observed a statistically significant increase of the predicted ( logarithmized ) IC50 values over the six time periods to each of the 11 bNAbs ( P ≤ 0 . 001 using the umbrella test and a significance threshold t = α/#tests = 0 . 05/22 = 0 . 0023 with Bonferroni correction for multiple testing ) . Thus , we could confirm the trend towards bNAb resistance [55–57] on a larger and more diverse data set . The predicted ( logarithmized ) IC50 values for the subtype B samples for all 11 bNAbs are provided in S3 Fig . Note that in order to avoid misleading data visualization , we present all the predicted values for all 11 bNAbs on the same y-scale , though the bNAbs differ in their neutralization strength . Though we find the last time periods as part of a significant trend in the data for PG9 , PG16 and PG128 , the partial p-values indicate rather a plateau distribution than a clear trend towards resistance in the last time periods ( see S6 Table ) . In addition , we predicted and analyzed the neutralization sensitivity of the non-B subtype samples ( 16 , 546 ) to the 11 bNAbs . A statistically significant trend towards resistance was observed for all considered bNAbs , but PG9 , PG16 , PG121 , PGT128 and NIH-4546 . In Fig 3 we show exemplarily the predicted ( logarithmized ) values for the bNAbs ( A ) 3BNC117 ( CD4bs ) , ( B ) PGT121 ( V3 loop ) , ( C ) 35O22 ( gp41/gp120 ) , and ( D ) PG16 ( V1/V2 loop ) ; see S4 Fig for all bNAbs and non-B subtype samples . While for the bNAb PGT128 there was no significant peak at all , the trend towards resistance to the bNAb PGT121 was not significant after Bonferroni correction for multiple testing . For PG9 , PG16 and NIH-4546 , we detected a significant peak in the data , but not in the last time period , which we however required to determine a trend towards resistance ( see Methods for details ) . The peak for NIH-4546 was slightly shifted ( in the fifth time period ) , whereas for PG9 and PG16 a significant peak was already detected in the first time period , that is , the HIV variants tend to become more susceptible in the last time period . Since there are no experimental data on HIV-1 resistance development trends to bNAbs for the non-B subtype population , we decided to first rule out the possibility of a confounder that might lead to the contrasting trend for PG9 and PG16 . Pfeifer et al . [63] discovered that there is a statistically significant bias in the neutralization susceptibility of HIV-1 variants to PG9 and PG16 depending on the coreceptor usage of the virus . For successful entry of the virus into the host cell , the glycoprotein gp120 has not only to bind to the CD4-receptor on the host cell , but also to a second chemokine receptor on the host cell that acts as co-factor ( coreceptor ) . The coreceptors mainly used by HIV-1 are CCR5 and CXCR4 . Depending on the coreceptor usage , the virus strain is referred to as R5- or X4-tropic , or dual-tropic if the virus can bind to both of these coreceptors , and X4-capable , if they are either dual-tropic or X4-tropic [64] . X4-capable viruses have been shown to be more resistant to PG9 and PG16 [63] . This means that PG9 and PG16 have an R5-bias , that is , they are better in neutralizing R5-tropic viruses than X4-capable viruses . By determining the coreceptor usage for all considered viral samples with the most widely used tool for genetic tropism testing , geno2pheno[coreceptor] [65] , we detected a stronger increasing ratio of R5- to X4-capable viruses over the time periods for the non-B than for the subtype B samples ( see Fig 3E and S2 Table ) . Thus , we might see an increase in neutralization susceptibility to PG9 and PG16 due to the relative increase of R5-tropic variants in the later time periods , since R5-tropic variants are more susceptible to PG9 and PG16 . With an analysis , analogous to Pfeifer et al . [63] , we observed an R5-bias of the bNAb PGT128 ( P = 0 . 00568 using a two-sided Fisher’s exact test , see also S3 Table ) . Fig 4 shows the relative number of resistant and susceptible HIV strains to PGT128 in comparison to PG9 , PG16 , VRC-PG04 and VRC01 . Data for VRC01 , VRC-PG04 , PG9 and PG16 was taken from Pfeifer et al . [63] . We additionally analyzed the association between coreceptor usage and neutralization sensitivity for all considered 11 bNAbs . As can be seen in S5 Fig , we could not detect other bNAbs with an R5-bias . For the bNAb PG16 , a resistance trend was only detected for the R5-tropic variants ( see Fig 3F ) . Note that sequences from the beginning of the HIV epidemic ( first two time periods ) were probably from patients having AIDS and not at early stage of HIV infection as nowadays . Since at early stage of clinical HIV infection usually R5-tropic viruses are predominant [66 , 67] , this might also explain the decrease of X4-capable variants in the database over time . The first time period contains also less samples than later time periods , which might influence the trend . We could detect a trend towards resistance for all 11 bNAbs regarding the subtype B HIV-variants ( 10/11 if Jonckheere-Terpstra test is used ) . For the non-B subtype population , we observed the trend for only 6 of the 11 bNAbs ( 5 of 11 if Jonckheere-Terpstra test is used ) . A summary of the findings and the corresponding p-values of both statistical tests can be found in S6 Table . In this study , we showed that neutralization sensitivity of new HIV-1 variants to broadly neutralizing antibodies ( bNAbs ) is predictable using neutralization information from existing neutralization assays . The credibility of the models were underlined by the finding that the prediction models learnt important binding sites for the bNAbs implicitly , without explicitly getting this type of information in the learning process . Hence , additional information such as structural binding site information is unlikely to boost the performance significantly . We increased the interpretability of the models , by offering the user more information on the prediction outcome in form of a motif logo where the logo displays the contribution of the pivotal residues of the test sequence to the prediction . In general , our method could be applied as a recommendation tool for bNAbs therapy , but it could already be used in planning clinical trials concerning bNAbs therapy to screen patients before those therapies are approved for clinical use . It is unquestioned , that an effective bNAb therapy will consist of a combination of bNAbs targeting distinct epitopes on the envelope spike to prevent the emergence of antibody resistance . To determine which and how many bNAbs to choose , several studies analyzed systematically combinations of different bNAbs [68 , 69] experimentally but also predicted the neutralization sensitivity using additive models . However , these prediction models need the neutralization sensitivity of the virus to the single bNAbs in the combination as input . Our learnt classifiers could be extended similarly to additive models that predict if or how effective a combination of bNAbs is requiring only the envelope sequence of the virus . Despite the good performance and biological relevance of our classifiers , the current models are not suited for a direct application in clinical settings . In the clinical setting , it is more tolerable to misclassify a sensitive HIV variant to a bNAb than misclassifying a resistant HIV-1 variant . While the area under the ROC curve was helpful in determining , if the classification task can be accomplished with our proposed methods and for comparison reasons , it is not the best approach to design the models for the final application setting due to the low average specificity for some of the bNAbs ( see S5 Table ) . In order to apply our models in the clinical setting , clinical data has to be analyzed instead of pseudovirus panel data . In addition , an appropriate false discovery rate has to be agreed on with the clinicians , for which the final models can be optimized for . This holds for any method used for this classification task . Apart from their potential use as recommendation tool , computational prediction models can in general be used to analyze the change in the neutralization sensitivity of HIV-1 over time . We could confirm previous results suggesting a trend towards antibody resistance in the subtype B population [55–57] . Moreover , we scaled up the analysis to the global HIV-1 population , showing that there is a general drift towards antibody resistance in the world-wide HIV-1 population for most of the bNAbs . These findings are relevant for the selection of suitable vaccine candidates; a combination of bNAbs is however still very potent in neutralizing HIV-1 [56] .
We used the IC50 titers of 11 different antibodies ( PG9 , PG16 , 35O22 , VRC01 , VRC-PG04 , 3BNC117 , NIH45-46 , PGT128 , PGT121 , 10-996 and 10-1074 ) for 115 to 220 HIV-1 isolates from three different neutralization assays [10 , 26 , 47] ( see S8 Table ) . For the bNAbs PG9 , PG16 , PGT121 and VRC01 neutralization information was available from two neutralization assays . Although the overlap of tested HIV-1 samples was quite high as well as the correlation of the corresponding IC50 titers , we did not merge the information from the two assays for these bNAbs . We represented each HIV-1 isolate by the amino acid sequence of its envelope glycoprotein from the Los Alamos HIV sequence database [48] . We excluded HIV-1 isolates for which no GenbankID was available , or the envelope sequence was shorter than 800 amino acids . Since the feature vector of each sample has to be of the same length for most of the kernels , we aligned the amino acid sequences with the HIValign tool from the Los Alamos HIV sequence database [48] . For the polynomial and Gaussian RBF kernel the amino acid sequences have to be transformed to a real-valued input . We used one-hot encoding to represent the sequence information for the polynomial kernel , i . e . , each amino acid ai , i ∈ {1 , …20} is transformed into a 20-dimensional vector , where only the i-th entry is 1 , and the others are 0 . For the Gaussian RBF kernel , we encoded the sequence information using physico-chemical properties ( RBF1 [70] and RBF2 [71] ) . In the classification task , the IC50 titers were converted to -1 if the IC50 value was above 50 μg/mL ( resistant ) , and otherwise to +1 ( susceptible ) similar to Doria-Rose et al . [47] . Due to their distribution , the IC50 values for the regression task were logarithmized . To test if l-mer string kernels ( such as the oligo kernel [52] or the weighted degree kernel with shifts ( WDKS ) [51] ) perform better than conventional kernels ( such as the polynomial or the Gaussian RBF kernel ) , we compared the performances of prediction models based on each of these kernels . The comparison was conducted by 10 runs of a 5-fold nested cross-validation using AUC and Pearson Correlation Coefficient as performance measure for the classification and regression task , respectively . The tested parameter range for each kernel is listed in S1 Table . To determine the best parameter setting for each bNAb prediction model , we performed an additional 5-fold cross-validation . Since in the nested cross-validation mainly small values of the width parameter ( 2σ2 ) led to high prediction performances , we further sampled the range between 0 and 3 for this parameter . We compared the final SVM classifiers based on the oligo kernel with random forests , an SVM using a linear kernel , a neural network , and a logistic regression with lasso regularization ( lasso ) . For the random forest , the neural network , and the lasso approach , the amino acid sequences were mapped to their index in the amino acid alphabet . For the linear kernel , the sequences have been encoded using the one-hot encoding approach , i . e . , each amino acid ai , i ∈ {1 , …20} is transformed into a 20-dimensional vector , where only the i-th entry is 1 , and the others are 0 . While the random forest approach can handle internally categorical variables with more than two factors , we created dummy features for each alignment position for the neural network and the lasso approach . We used the R package randomForest [72] setting the number of variables randomly sampled as candidates at each split ( mtry ) to the square root of the number of features in the model and the numbers of tree to grow ( ntree ) to 500 . For the neural network , we used the R package neuralnet [73] , we used one layer and set the number of hidden layers to the square root of the number of features . To build the logistic regression models , we used the R package glmnet [74] where we used lasso as regularization ( α = 1 ) and tuned the lambda parameter in an internal cross-validation . For the linear kernel , we used the R package kernlab [75] setting the kernel to vanilladot using the default cost parameter C . The performance was assessed on 10 runs of stratified 5-fold cross-validation . We did not compare the performance over a nested cross-validation iterating over different hyperparameters for the models , due to the infinite range of possibilities . We used the R package mlr [76] to compare all the methods . A kernel k ( x , x′ ) can be considered as a similarity function between instances x and x′ . The oligo kernel computes k ( x , x′ ) between two sequences x and x′ of same length L by comparing the co-occurrences of their substrings ( oligomer ) of length l with 1 ≤ l ≤ L within a particular distance ( width parameter σ2 ) . Therefore , the occurrence of a particular l-mer in a sequence x ( denoted as xω ) is encoded by the so-called oligo function μ μω ( t ) =∑p∈xωexp ( −12σ2 ( t−p ) 2 ) ( 1 ) with the continuous position variable t ∈ [1 , L] and σ2 controlling the positional uncertainty . As described in [52] , the corresponding learnt weight of the classifier for each oligomer ωat each position t can be retrieved by | w ω ( t ) | = | ∑ i = 1 N α i y i μ ω i ( t ) | , ( 2 ) where i ∈ {1 , … , N} denotes the i-th training sample with αi ≥ 0 and yi ∈ {−1 , 1} being the learnt weight and classification label of the i-th sample , and with μ ω i ( t ) being the oligo function of l-mer ω of the training sequence i at position t . Considering the weights of each oligomer for the test sequence , there exists only one oligomer ω containing the actual residue as starting point whose contribution is calculated as S ω * ( t ) = ∑ i = 1 N α i y i < μ ω i ( t ) , μ ω * ( t ) > , ( 3 ) with μ ω * being the oligo function of l-mer ω of the test sequence . For l-mers > 1 the computed contribution is assigned to all amino acids of the oligomer . To visualize the motif logos we used Weblogo 3 . 0 [77] . We used all available envelope sequences from the Los Alamos HIV sequence database [48] ( 37 , 137 ) , except the sequences that the prediction models were built on and those that were too short , resulting in 35 , 524 envelope sequences . For 1586 sequences no date was given , and thus these sequences were excluded as well , resulting in 33 , 938 considered viral envelope sequences . Before predicting the IC50 value for each test sequence , the sequences were aligned to the data sets using profile-to-profile MUSCLE alignment [78] with the Ugene tool [79] . Instead of predicting the IC50 value , the regression models were trained to predict the logarithmized IC50 value . To identify a drift towards resistance , we performed a permutation test for umbrella alternatives [58] on the predicted ( logarithmized ) IC50 values grouped by the six time periods . We applied the umbrella test according to Basso et al . [58 , 62] with the provided R code . The umbrella test is a generalization of the Jonckheere-Terpstra test , testing for a peak instead of a monotone trend . A significant peak in the last time period was considered as indicator for an increasing trend in IC50 values and thus , a trend towards bNAb resistance . To predict the coreceptor usage , we used the well established prediction tool geno2pheno[coreceptor] [65] . The prediction tool uses a linear support vector machine to predict whether a sequence is from a X4-capable or an R5-tropic virus , only based on the V3 loop sequence of the viral envelope sequence . For each V3 sequence , geno2pheno[coreceptor] provides the false-positive rate ( FPR ) , which is a measure for the confidence of the prediction . geno2pheno[coreceptor] reports the minimal FPR threshold at which the sequence would be classified as X4-capable . For the manuscript , we used an FPR cutoff of 10% to determine X4-capable ( ≤ 10% ) and R5-tropic viruses ( > 10% ) as recommended by the European Consensus Group on clinical management of HIV-1 tropism testing [80] . Since there are also reasons for other cutoff choices , we additionally provide the results for the FPR cutoffs according to the German and Austrian treatment guidelines ( ≤ 5%: X4-capable; ≥ 15%: R5-tropic ) in the Supporting Information . We used the prediction tool geno2pheno[coreceptor] [65] to determine the coreceptor usage of the 33 , 938 viral isolates from the Los Alamos HIV sequence database [48] . According to the prediction tool , we excluded in total 545 sequences due to warnings regarding the alignment quality and due to warnings regarding the V3 loop quality ( alignment score ≥ 95th percentile ) . For this analysis , we used all available sequences from the CATNAP tool [81] , retrieved on 2016-08-10 . Since we used a neutralization sensitivity cutoff of 50 μg/mL to determine resistance and susceptibility , all sequences , whose neutralization sensitivity were only given as a cutoff less than 50 μg/mL were excluded . In addition , we excluded two sequences due to poor V3 alignment quality . Coreceptor usage was determined using the prediction tool geno2pheno[coreceptor] [65] . To test whether the sensitivity to an antibody is significantly different with regard to coreceptor usage , we performed a two-sided Fisher’s exact test for the two-by-two contingency tables with resistant and susceptible as the row label and X4-capable/R5-tropic as the column label using significance level = 0 . 05 with the null hypothesis that there is no difference . The prediction and analysis of the neutralization sensitivity were implemented mainly in R [82] , version 3 . 2 . 1 ( 2015-06-18 ) and the R package kernlab [75] . The oligo kernels were computed using a customized version of the Shogun-Toolbox [83] ( version 2 . 0 . 0 ) . To visualize the motif logos we used Weblogo 3 . 0 [77] . In S8 Table we provide the virus names that we considered for the prediction models as well as the study ID of the neutralization assay . With this , the corresponding neutralization data can be retrieved from CATNAP [81] . For the trend analysis , we provide the accession numbers of each considered HIV-1 variant in the Los Alamos HIV sequence database [48] in S9 Table . At https://github . com/annahake/g2p-bnab , we additionally provide the computed kernels for the final models as well as the resampling instance for the 10 runs of stratified 5-fold cross-validation . As mentioned in the Conclusion , the final models are so far not adapted for clinical usage . | Several sequence-based approaches exist to predict the epitope of broadly neutralizing antibodies ( bNAbs ) against HIV based on the correlation between variation in the viral sequence and neutralization response to the antibody . Though the potential epitope sites can be used to predict the neutralization response , the methods are not optimized for the task , using additional structural information , additional preselection steps to identify the epitope sites , and assuming independence and/or only linear relationship between the potential sites and the neutralization response . To model also the neutralization response to bNAbs with more complex binding sites , including for example several non-consecutive residues or accompanying conformational changes , we used non-linear , multivariate machine learning techniques . Though we used only the viral sequence information , the models learnt the corresponding binding sites of the bNAbs . In general only few residues were learnt to be responsible for a change in neutralization response , which can additionally reduce the sequencing cost for application in clinical routine . We propose our tailored models to aid the patient selection process for current clinical trials for bNAb immunotherapy , but also as a basis to predict the best combinations of bNAbs , which will be required for routine clinical practice in the future . | [
"Abstract",
"Introduction",
"Results",
"and",
"discussion",
"Materials",
"and",
"methods"
] | [
"chemical",
"neutralization",
"medicine",
"and",
"health",
"sciences",
"immune",
"physiology",
"pathology",
"and",
"laboratory",
"medicine",
"pathogens",
"immunology",
"microbiology",
"retroviruses",
"viruses",
"immunodeficiency",
"viruses",
"mathematics",
"rna",
"viruses",
"forecasting",
"statistics",
"(mathematics)",
"sequence",
"motif",
"analysis",
"coreceptors",
"antibodies",
"research",
"and",
"analysis",
"methods",
"sequence",
"analysis",
"immune",
"system",
"proteins",
"bioinformatics",
"proteins",
"medical",
"microbiology",
"hiv",
"mathematical",
"and",
"statistical",
"techniques",
"microbial",
"pathogens",
"biological",
"databases",
"chemistry",
"hiv-1",
"biochemistry",
"signal",
"transduction",
"sequence",
"databases",
"cell",
"biology",
"viral",
"pathogens",
"database",
"and",
"informatics",
"methods",
"physiology",
"biology",
"and",
"life",
"sciences",
"chemical",
"reactions",
"physical",
"sciences",
"lentivirus",
"statistical",
"methods",
"organisms"
] | 2017 | Prediction of HIV-1 sensitivity to broadly neutralizing antibodies shows a trend towards resistance over time |
Telomerase reverse transcriptase ( TERT ) and telomerase RNA ( TR ) represent the enzymatically active components of telomerase . In the complex , TR provides the template for the addition of telomeric repeats to telomeres , a protective structure at the end of linear chromosomes . Human TR with a mutation in the template region has been previously shown to inhibit proliferation of cancer cells in vitro . In this report , we examined the effects of a mutation in the template of a virus encoded TR ( vTR ) on herpesvirus-induced tumorigenesis in vivo . For this purpose , we used the oncogenic avian herpesvirus Marek's disease virus ( MDV ) as a natural virus-host model for lymphomagenesis . We generated recombinant MDV in which the vTR template sequence was mutated from AATCCCAATC to ATATATATAT ( vAU5 ) by two-step Red-mediated mutagenesis . Recombinant viruses harboring the template mutation replicated with kinetics comparable to parental and revertant viruses in vitro . However , mutation of the vTR template sequence completely abrogated virus-induced tumor formation in vivo , although the virus was able to undergo low-level lytic replication . To confirm that the absence of tumors was dependent on the presence of mutant vTR in the telomerase complex , a second mutation was introduced in vAU5 that targeted the P6 . 1 stem loop , a conserved region essential for vTR-TERT interaction . Absence of vTR-AU5 from the telomerase complex restored virus-induced lymphoma formation . To test if the attenuated vAU5 could be used as an effective vaccine against MDV , we performed vaccination-challenge studies and determined that vaccination with vAU5 completely protected chickens from lethal challenge with highly virulent MDV . Taken together , our results demonstrate 1 ) that mutation of the vTR template sequence can completely abrogate virus-induced tumorigenesis , likely by the inhibition of cancer cell proliferation , and 2 ) that this strategy could be used to generate novel vaccine candidates against virus-induced lymphoma .
Telomerase is a multi-component ribonucleoprotein complex that governs the maintenance of telomeres , protein-associated hexameric sequence repeats at the end of linear chromosomes , and ensures chromosomal integrity and cellular survival [1] , [2] . The telomerase complex consists of two core components , telomerase reverse transcriptase ( TERT ) and telomerase RNA ( TR ) . In the complex , TR serves as the template for TERT , which catalyzes the addition of telomeric repeats ( TTAGGG ) n at chromosome ends [3] . Vertebrate TRs exhibit a universally conserved secondary structure comprised of four structural domains ( Fig . 1 ) : the pseudoknot ( core ) domain containing the template sequence in conserved region ( CR ) 1 ( CR1 ) , the CR4 and CR5 domains with a highly conserved stem-loop structure ( CR4-5 ) , the H/ACA box domain , and the CR7 domain [4] . CR1 encodes the template sequence that is utilized for the extension of the telomeric repeats , while the CR4-5 domain contributes to the processivity of telomerase and is essential for stable assembly with TERT . The H/ACA box and CR7 domains confer TR stability [4]–[6] . Telomerase activity is absent in most somatic cells , but commonly up-regulated in rapidly dividing cells including transformed cells [7] . Consistent with this observation , telomerase activity is significantly elevated in over 85% of human cancers and over 70% of immortalized human cell lines [8] . The absence of telomerase activity often leads to progressive telomere shortening resulting in cellular senescence and irreversible cell cycle arrest [9] . Several tumor-inducing viruses have evolved strategies to evade or subvert mechanisms controlling cellular senescence , mainly via the up-regulation of TERT , which is generally the limiting factor for telomerase activity [10]–[13] . It has been suggested that up-regulation of TERT expression and , consequently , increased telomerase activity ensures the proliferative potential of persistently infected cells . One of the most efficient viruses with respect to induction of fatal tumors is Marek's disease virus ( MDV ) . MDV is a lymphotropic herpesvirus that causes a well-described syndrome , Marek's disease ( MD ) , in chickens . MD is characterized by neurological disorders , immune suppression , and malignant T cell lymphomas [14] . The rapid onset of lymphomas developing within 2 to 3 weeks post-infection ( p . i . ) and high tumor-induced mortalities of 90–100% in susceptible chickens make MDV-induced transformation an ideal model to study virus-induced tumorigenesis in a natural virus-host setting [15] . A number of MDV-encoded genes have been shown to be involved in MDV-induced transformation . The major MDV oncogene , meq , encodes a basic leucine zipper ( bZIP ) transcription factor similar to the cellular homologues c-Jun , c-Fos , and c-Myc . Meq dimerizes with other bZIP transcription factors and modulates expression of both cellular and viral genes [16] , [17] . MDV also encodes other genes products and sequence elements , which perform auxiliary functions in transformation [18] . One such element is a TR homologue termed viral TR ( vTR ) that shares 88% sequence identity with chicken TR ( chTR ) [19] . The high sequence homology suggests vTR was likely acquired from the chicken genome during virus-host co-evolution . Compared to its cellular counterpart , chTR , interaction of vTR with TERT results in higher telomerase processivity [20] , [21] . It was shown that vTR contributes to the rapid onset of lymphoma formation by serving as a template for TERT , but it also has functions that are independent of the telomerase complex . It is predominantly the telomerase-independent functions of vTR that are responsible for tumor progression and dissemination [21] , [22] . In vitro experiments demonstrated that mutations in the template sequence within CR1 of human and mouse TR can result in telomere instabilities , aberrant chromosome separation and segregation , and ultimately apoptosis [23] , [24] . TRs with a mutated template can induce unique checkpoint responses that are different from DNA damage or loss-of-telomerase responses , even at low mutant TR expression levels and in the presence of wild-type TR . In addition , pro-apoptotic effects were also shown for TRs harboring mutant template or oligonucleotides specifying mutant template sequences and such molecules are discussed as anti-tumor therapeutics in different types of cancers [23] , [25] , [26] . Here , we investigated the effect of a mutant vTR template sequence ( AU5 ) on the tumor-promoting capacity of a highly oncogenic avian herpesvirus in its natural host . Mutation of the template sequence of MDV-encoded vTR completely abrogated virus-induced tumor formation in chickens . Introduction of a second mutation in the stem loop ( CR4-5 ) region that abolishes a functional interaction of vTR with TERT restored lymphomagenesis , confirming that the abrogation of tumorigenesis shown for the mutant virus is dependent on telomerase activity through interaction of mutant vTR and TERT . In vaccination-challenge studies , the virus expressing mutant template protected chickens from lethal challenge with a very virulent MDV strain .
TRs harboring mutations in the template sequence were previously shown to result in telomere instabilities , aberrant chromosome separation and segregation , and ultimately apoptosis in mammalian cells in vitro [23] . Led by these previous observations , we hypothesized that expression of vTR encoding a mutated template sequence ( AATCCCAATC to ATATATATAT ) , termed AU5 , could have an effect on avian cancer cells that is similar to that described previously for mammalian cells [23] , [24] . In order to test our hypothesis , we first screened a number of avian primary cells and permanent cancer cell lines to determine the optimal system that would provide sufficient levels of telomerase activity . We performed telomere repeat amplification protocol ( TRAP ) assays to detect telomerase activity in primary chicken embryo cell ( CEC ) cultures , the chicken fibroblast cell line DF-1 [27] , and the quail cancer cell line QT35 [28] . CEC and DF-1 cells did not exhibit telomerase activity , while the QT35 cancer cell had high telomerase activity as evidenced by the presence of numerous TRAP products ( Fig . 2A ) . A DF-1 cell line stably expressing TERT showed some telomerase activity , suggesting that TERT was the limiting factor for telomerase activity in this cell line . Based on the results , we used a previously established QT35 cancer cell line that allowed tetracycline-inducible expression [29] of vTR or vTR-AU5 ( AU5 ) . During the establishment of cell lines we observed that even un-induced AU5 cell lines replicated markedly slower . From many initial clones , only a single monoclonal AU5 cell line could be established , suggesting a strong selection against leaky AU5 expression . This effect has been previously observed during the development of mammalian cell lines expressing TR template mutants [23] . Therefore , polyclonal vTR and AU5 cell lines were used to determine the effect of AU5 expression on cancer cell proliferation . RT-qPCR analysis of polyclonal cells , confirmed leaky expression of the constructs and that vTR and AU5 expression could be increased by more than 300-fold upon induction with doxycycline after 5 days of treatment ( Fig . 2B ) . To determine if AU5 inhibits cancer cell proliferation , we analyzed colony formation by measuring confluency over 31 days in the presence or absence of doxycycline . Constitutive or induced expression of wild-type vTR resulted in enhanced proliferation when compared cells harboring the vector control as described previously [21] . In contrast , cell lines harboring AU5 exhibited a significant growth defect when compared to vTR and control cell lines ( Fig . 2C ) . Increased expression of AU5 following induction resulted in only slightly reduced cell proliferation when compared with non-induced cells , suggesting that expression of only low levels of AU5 are sufficient to reduce growth of the QT35 cancer cell line . We concluded that our results are consistent with those of TR over-expression in human and murine cancer cells [23] , [24] and show that expression of the MDV vTR can help stabilize and/or promote growth , while mutation of the template sequence significantly impairs proliferation of the avian QT35 cancer cell line . Since mutation of the template sequence of vTR resulted in decreased proliferation of QT35 cancer cells ( Fig . 1C ) , we hypothesized that the mutation in the context of virus infection may have an effect on MDV replication and tumorigenesis in vivo . Therefore , we mutated the template sequence of vTR ( AU5 ) in pRB-1B , an infectious bacterial artificial chromosome ( BAC ) clone of the highly oncogenic RB-1B MDV strain using two-step Red-mediated recombination [30] , [31] . Two rounds of mutagenesis allowed the desired alteration of both copies of the diploid vTR gene within the MDV genome , and transfection of the recombinant BAC clone into CEC resulted in the reconstitution of the vTR template mutant virus ( vAU5 ) . Furthermore , a revertant clone ( AU5rev ) was generated in which the wild-type template sequence was restored in the mutant vAU5 . Following virus reconstitution , we performed plaque size assays and multi-step growth kinetics in CEC that revealed that the growth properties of vAU5 were indistinguishable from those of parental ( vRB-1B ) and revertant ( vAU5rev ) viruses ( Fig . 3A–D ) . Next , we determined if expression of AU5 had an effect on MDV replication , disease , and tumor incidence in vivo . In two independent experiments , we infected 1-day-old P2a chickens with vRB-1B , vAU5 , or vAU5rev and monitored virus levels in the blood using qPCR assays until 28 days post infection ( dpi ) . vAU5 replication was significantly impaired when compared to parental and revertant viruses , indicating that the number of infected B and T cells is reduced ( Fig . 4B and D ) . Consistent with the reduction of viremia , none of the chickens infected with vAU5 developed tumors in two independent experiments ( 0/10; 0/18 ) over the course of 13 weeks while parental ( vRB-1B ) or revertant ( vAU5rev ) viruses induced lymphomas in 92–100% of infected animals ( Fig . 4A and C ) . We concluded from our data that expression of vTR harboring the AU5 mutation by MDV can completely abrogate virus-induced tumorigenesis in highly susceptible chickens , most likely by the elimination of MDV-infected and/or transformed cells by apoptosis . We previously demonstrated that a mutation within the vTR P6 . 1 stem-loop can prevent incorporation of vTR into the telomerase complex and abolish enzymatic activity and telomere elongation [22] . To confirm that the absence of lymphoma in vAU5-infected animals was dependent on the presence of AU5 in the telomerase complex , we constructed mutant viruses in which the AU5 and P6 . 1 mutations were introduced into vTR either individually or together . Revertant viruses of each mutation were also generated . All constructed viruses replicated with kinetics comparable to those of parental and revertant viruses in vitro ( Fig . 5 ) . Upon infection of chickens with the recombinant viruses , qPCR analysis revealed that insertion of the P6 . 1 mutation into vAU5 ( vAU5+P6 . 1mut ) restored lytic virus replication to levels comparable to those of parental vRB-1B , while mutant virus only harboring the AU5 mutation ( vAU5+P6 . 1rev ) was significantly impaired in replication ( Fig . 6A ) . Like vAU5 , vAU5+P6 . 1rev did not induce tumors in any of the infected chickens ( 0/19 ) ( Fig . 6B ) . Two of the 19 chickens ( 11% ) died over the course of the 13 week experiment , which was likely due to immunosuppression and generalized wasting , common characteristics of MD and observed in earlier reports using viruses that are unable to express vTR [21] . Viruses that contained the P6 . 1 and the AU5 mutation ( vAU5+P6 . 1mut ) induced lymphomas in 100% of infected animals . Furthermore , vP6 . 1 , parental and complete revertant viruses caused tumors in all animals infected with the respective viruses . From the results we concluded that abrogation of lymphoma formation after infection with an MDV specifying the AU5 template mutation is indeed dependent on the interaction of template mutant vTR with TERT and has no effect if it is not incorporated into the telomerase complex . Since MDV harboring the template mutant vTR did not induce tumors but still replicated in chickens , we addressed the question whether vAU5 could induce a robust enough immune response to serve as a vaccine . Groups of 1-day-old P2a ( highly susceptible to MD ) and N2a ( partially resistant to MD ) chickens were inoculated with diluent , vAU5 , or the widely used , commercial vaccine strain CVI988 . Vaccinated chickens were challenged 10 days later with the very virulent RB-1B MDV strain . Chickens receiving the diluent developed tumors with expected frequencies of 100% in the P2a chickens and 79% in the N2a chickens after 13 weeks ( Fig . 7A and B ) [32] . vAU5 vaccinated N2a chickens were completely protected from lethal challenge , while 7% of the animals vaccinated with the commercial CVI988 vaccine strain developed disease with a protective index of 91% . In P2a chickens that are highly susceptible to MD , both vAU5 and CVI988 efficiently induced protection against challenge infection , as only 1 animal in each group developed disease . The protective index of vAU5 and CVI988 in P2a animals was 93% and 92% , respectively . These results suggest that mutation of the template region of vTR in a virulent MDV can serve as a strategy to induce protection against virus-induced lymphomas .
We here report on effects of a mutation in the template sequence ( CR1 ) of vTR encoded by MDV on virus replication and tumorigenesis in a natural virus-host model . Mutation of the vTR template sequence from AATCCCAATC to ATATATATAT ( AU5 ) resulted in decreased proliferation of the QT35 avian cancer cell line ( Fig . 2 ) , as had been described for TR in mammalian cells [23] , [24] . Introduction of the template sequence mutation in vTR in the context of the viral genome and infection of MD-susceptible chickens with mutant virus ( vAU5 ) resulted in complete absence of tumors and low-level viral replication in vivo ( Fig . 4 ) . Secondary mutation of the vTR stem-loop sequence ( P6 . 1 ) , abolishing the interaction of mutant vTR with TERT , restored virus-induced tumorigenesis ( Fig . 6 ) , thus showing that vTR-TERT interaction and functional telomerase activity is required for the anti-tumorigenic effects of the mutant template sequence in a viral background . Vaccination with MDV harboring the vTR template mutation not only abrogated herpesvirus-induced tumorigenesis , but also protected chickens from a lethal challenge with a very virulent MDV strain . We surmise that the reduced proliferation of QT35 expressing vTR with a template sequence mutation , as well as the absence of tumors and greatly reduced lytic replication in chickens are both caused by the incorporation of mutant telomeric repeat sequences into host telomeres of infected cells , which eventually leads to telomere crisis and apoptosis ( Fig . 8 ) . This sequence of events has been shown previously in other mammalian systems in vitro , where even low levels of mutant TR induced a unique checkpoint response resulting in telomere instabilities , aberrant chromosome separation and segregation , and apoptosis [23] , [24] . In addition , the pro-apoptotic effect of TRs harboring mutant templates or oligonucleotides specifying mutant template sequences has also been shown [23] , [25] , [26] . It is interesting to note that the QT35 cancer cell line was previously shown to maintain MDV in a latent state [33] and that the cells express MDV vTR at very low levels ( Fig . 2B ) . Despite the expression of endogenous quail TR and MDV vTR , AU5 expression had a negative effect on the replication of QT35 cancer cells . Consequently , induced over-expression of wild-type vTR led to increased proliferation of QT35 cancer cells , further lending support to the interpretation that vTR performs an important function in the early maintenance of transformed cells [21] . Induced expression of the AU5 sequence leading to incorporation of mutant template sequences significantly reduced proliferation presumably by inducing apoptosis , again consistent with previous studies on mammalian TRs [23] . vTR was previously shown to contribute to MDV-induced lymphomagenesis . Deletion of vTR in the MDV genome resulted in significantly reduced tumor incidences but the mutation did not affect virus replication in vivo [21] . Mutation of the vTR template region ( AU5 ) in MDV , however , completely abrogated tumorigenesis and reduced viral loads in infected animals , likely via inhibition of cancer cell replication through induction of apoptosis . vTR has at least two functions during lymphomagenesis , one that is dependent and one that is independent of telomerase activity . The telomerase-dependent function plays an important role in the early onset of disease but is dispensable for tumorigenesis . This conclusion is supported by studies showing that MDV harboring a mutated vTR incapable of interaction with TERT ( P6 . 1mut ) can still induce tumors in chickens , albeit resulting in a delayed onset of tumor formation [22] . vTR functions that are independent of its presence in the telomerase complex seem to be important for lymphomagenesis but are poorly understood . We utilized this previously published mutation to determine if the induction of apoptosis and abrogation of tumorigenesis is dependent on the incorporation of AU5 into the telomerase complex . While MDV harboring AU5 vTR are incapable of inducing tumors , mutation in the vTR-TERT interaction domain ( vAU5+P6 . 1mut ) in vAU5 completely restored tumorigenesis . We therefore concluded that restored ability of vAU5+P6 . 1mut to induce tumors is presumably caused by the inability of the telomerase to incorporate mutant telomeric repeats at the ends of host chromosomes and cause telomere crisis and apoptosis; hence the pro-oncogenic functions of vTR that are independent of the telomerase complex prevail . Our results demonstrate that vAU5+P6 . 1mut can efficiently cause lymphoma , which confirms that vTR has tumor-promoting functions independent of the telomerase complex and that they are mediated by a vTR domain outside of the template region . The fact that vAU5 was unable to induce tumors in highly susceptible P2a chickens suggested that it could serve as a potential vaccine against MD . Vaccination with vAU5 protected chickens from lethal challenge infection with the very virulent MDV strain RB-1B at least as efficiently as the commercial vaccine strain CVI988/Rispens that is commonly used in the field [34] . However , the residual mortality observed in some chickens has to be clarified to ensure the safety of the vaccine candidate . A similar phenomenon of low levels of mortality in highly susceptible birds , similar to the one observed here , was also observed with MDV mutants in which the major oncoprotein of MDV , Meq , was absent . Infection with meq deletion viruses did not cause tumors [35] , but severe lymphoid atrophy and immunosuppression was evident [36] . Likely , a combination of vTR template mutation with other modifications in the MDV genome targeting genes important for replication could therefore increase the safety of the vaccine and prevent lymphoid atrophy . Here , we suggest a new strategy that could be applied to the next generation of MD vaccines , which will certainly be needed because recently isolated MDV strains are capable of evading immune protection provided by current vaccines [15] , [37] , [38] .
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 Committee on the Ethics of Animal Experiments of Cornell University ( permit number 2002-0085 and 2008-0018 ) . The animal care facilities and programs of Cornell University meet the requirements of the law ( 89–544 , 91–579 , 94–276 ) and NIH regulations on laboratory animals , and are in compliance with the Animal Welfare Act , PL 279 . All experimental procedures were in compliance with approval of Cornell University's Institutional Animal Care and Use Committee ( IACUC ) and all efforts were made to minimize suffering . Recombinant viruses were generated by two-step Red-mediated recombination as previously described [30] , [31] . Primers used for construction of template sequence ( CR1 ) AU5 mutants ( vAU5 ) and revertants ( vAU5rev ) are shown in Table 1 . Primers used for construction of the P6 . 1 vTR-TERT interaction domain mutants and revertants have previously been published [22] . CEC cultures were prepared from 10-day-old specific-pathogen-free ( SPF ) embryos using standard methods [39] . Recombinant viruses were reconstituted from purified BAC DNA in CEC cultures using CaPO4 transfection [40] . The loxP flanked mini-F sequences within the infectious clones were removed by co-transfection with a Cre recombinase expression vector ( pCAGGS-NLS/Cre ) as previously described and screened via analytical PCR [30] . Virus propagation , plaque area measurements and multi-step growth kinetics were also performed as described previously [41] . SPF P2a ( MHC haplotype B19B19 ) or N2a ( MHC haplotype B21B21 ) chickens were obtained from departmental flocks and housed in poultry isolation units . Chickens were inoculated with 1 , 000 or 2 , 000 plaque forming units ( PFU ) of virus by intra-abdominal injection and evaluated for symptoms of MD on a daily basis . Necropsies were performed on chickens showing clinical signs of MD , as well as all remaining chickens at the termination of the experiment . DNA was extracted from whole blood of eight chickens for each group randomly selected prior to the experiment and MDV genomic copies were determined by qPCR assays [41] . Briefly , MDV DNA copy numbers were detected using primers and probe specific for the MDV infected cell protein 4 ( ICP4 ) locus that were normalized to cellular genome copies of chicken inducible nitric oxide synthase ( iNOS ) . Tet-on constructs were generated by digestion of the pCMS-vTR and pCMS-vTR-AU5 constructs previously described [22] with EcoRI and XbaI . Resulting vTR or AU5 fragments were then cloned into the pcDNA4/TO/myc-his vector ( Invitrogen , Carlsbad , CA ) to generate pcDNA4/TO-vTR and pcDNA4/TO-AU5 , respectively . Inducible cell lines were generated based on QT35TR19 , a previously described Tet-inducible QT35 cancer cell line ( kindly provided by Karel A . Schat , Cornell University ) and maintained as described previously [29] . To generate control , vTR , and AU5 expressing cell lines , QT35TR19 cells were transfected with pcDNA4/TO ( empty vector ) , pcDNA4/TO-vTR , or pcDNA4/TO-AU5 using Lipofectamine2000 ( Invitrogen , Carlsbad , CA ) following the manufacturer's instructions . Monoclonal and polyclonal cell lines were selected with 5 µg/ml blasticidin and 500 µg/ml zeocin ( Invitrogen ) . All cell lines were used between passage 5 and 10 in cell proliferation and RT-qPCR assays . Proliferation of Tet-inducible cell lines was evaluated as previously described [23] . Briefly , 2×103 cells of each cell line were seeded into 35 mm dishes in triplicate and maintained in media with or without 1 µg/ml doxycycline with 2/3 media changed every 4–5 days . After 31 days , cells were fixed with 90% ice-cold acetone and stained with 1% crystal violet in 50% methanol . Percent confluency was determined using NIH ImageJ software by calculating the total area on the plates covered by cell colonies over the total area of the plate . The average % confluency was determined from three independent experiments . One-thousand vTR , AU5 , or empty vector Tet-inducible cells were treated with or without 1 µg/ml doxycycline in triplicate . After 3 and 5 days , total RNA was prepared using RNA STAT 60 as described previously [42] . Reverse transcription was performed using the ThermoScriptTM RT-PCR system ( Invitrogen , Carlsbad , CA ) with random hexamer oligonucleotides according to manufacturer's instructions . Copies of vTR cDNA were determined by qPCR assays using the TaqMan Fast Universal Master Mix system ( Applied Biosystems , Inc . ) according to manufacturer's instructions and performed in an ABI Prism 7500 Fast Real-Time PCR System ( Applied Biosystems , Inc . ) . Results were analyzed with the Sequence Detection Systems version V2 . 0 . 3 software using the comparative Ct method ( 2−ΔΔCt ) of relative quantification . Primers and probe for the detection of MDV vTR and quail glyceraldehyde-3-phosphate dehydrogenase ( GAPDH ) that served as an endogenous control have been described previously [43] , [44] . Significant differences in % confluency assays and MDV replication using qPCR assays were determined using Student's t test or Tukey-Kramer comparison of means . | Telomerase facilitates telomere maintenance and consists of two major components: the catalytic subunit telomerase reverse transcriptase ( TERT ) and telomerase RNA ( TR ) that provides the template for the addition of telomeric repeats to telomere ends . Expression of TRs with a mutation in the template sequence can result in telomere instability , cell cycle arrest and apoptosis in mammalian cells . Here , we introduced a template mutation in a TR encoded by the highly oncogenic avian herpesvirus Marek's disease virus ( MDV ) to evaluate this mechanism in a natural virus-host model for virus-induced tumorigenesis . Expression of the mutant viral telomerase RNA ( vTR ) by MDV allowed virus replication in telomerase-deficient cells , but completely abrogated MDV-induced lymphomagenesis in vivo in a telomerase-dependent manner . Infection with MDV harboring the template mutation in vTR not only abrogated herpesvirus-induced tumorigenesis , but also protected chickens from lethal challenge with a very virulent MDV strain . We provide the first in vivo evidence that a TR with a template mutation expressed by a herpesvirus can be used to prevent herpesvirus-induced tumorigenesis , an approach that could be used for the development of the next generation vaccines against MDV and possibly also other lymphotropic herpesviruses . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"medicine",
"infectious",
"diseases"
] | 2011 | Herpesvirus Telomerase RNA (vTR) with a Mutated Template Sequence Abrogates Herpesvirus-Induced Lymphomagenesis |
Plasmodium vivax is the most widely distributed human malaria parasite with 2 . 9 billion people living in endemic areas . Despite intensive malaria control efforts , the proportion of cases attributed to P . vivax is increasing in many countries . Genetic analyses of the parasite population and its dynamics could provide an assessment of the efficacy of control efforts , but , unfortunately , these studies are limited in P . vivax by the lack of informative markers and high-throughput genotyping methods . We developed a sequencing-based assay to simultaneously genotype more than 100 SNPs and applied this approach to ~500 P . vivax-infected individuals recruited across nine locations in Cambodia between 2004 and 2013 . Our analyses showed that the vast majority of infections are polyclonal ( 92% ) and that P . vivax displays high genetic diversity in Cambodia without apparent geographic stratification . Interestingly , our analyses also revealed that the proportion of monoclonal infections significantly increased between 2004 and 2013 , possibly suggesting that malaria control strategies in Cambodia may be successfully affecting the parasite population . Our findings demonstrate that this high-throughput genotyping assay is efficient in characterizing P . vivax diversity and can provide valuable insights to assess the efficacy of malaria elimination programs or to monitor the spread of specific parasites .
Plasmodium vivax is the most widely distributed human malaria parasite , threatening almost 40% of the world population [1] . It is the major cause of malaria outside of Africa , with up to 390 million clinical infections each year , and is responsible for long-term chronic illness that has dramatic consequences for the health and economy of endemic regions . Despite the implementation of extensive control efforts , the proportion of cases attributed to P . vivax has significantly increased over time compared to P . falciparum [2] . For example , in Cambodia the malaria control activities in the last ten years–extended access to insecticide treated bed nets , distribution of rapid diagnostic tests ( RDTs ) to health facilities , switch to dihydroartemisinin-piperaquine as first line therapy for treatment of uncomplicated malaria–have led to a 50% decrease in malaria cases ( from 113 , 855 cases in 2004 to 56 , 271 cases according to data from the Ministry of Health ) . However , these measures disproportionally affected falciparum malaria: for the first time in 2011 , the number of P . vivax infections ( 42 , 901 ) became higher than the number of P . falciparum infections ( 33 , 326 ) and by 2014 , P . vivax was detectable in 78% of the malaria cases ( 46% of mono-infections and 32% of mixed infections ) . In order to effectively design elimination strategies to combat P . vivax , it is essential that we better understand how the parasite population is changing in response to control measures . This is particularly important for P . vivax since our inability to culture the parasite in vitro [3] complicates measuring the efficacy of antimalarial drugs in the population ( though ex vivo studies can partially overcome this limitation [4–6] ) . Unfortunately , the lack of an in vitro culture system also limits the amount of genetic and genomic data available for P . vivax . Most population genetic studies in P . vivax have relied on small numbers of microsatellite markers ( e . g . , see [7–9] ) or single nucleotide polymorphisms ( SNPs ) [10 , 11] or targeted a few protein coding genes [12] . These studies therefore have limited power to identify population structure ( that typically requires many unlinked markers ) or to detect the presence of multiple , genetically-different parasites in a single infection ( later referred to as complexity of infection or COI ) . Whole genome sequencing studies have the potential to circumvent these limitations but i ) are prohibitively expensive for analyzing hundreds of samples and ii ) are hindered by the presence of the patient DNA and require processing the samples before sequencing ( e . g . , using CF11 columns [13] or sequence capture [14] ) . We describe here a novel genotyping assay that provides sequence information from over 100 selected polymorphic loci and allows robust assessments of P . vivax population genetic diversity and COI . This technique enables high-throughput genotyping of parasite DNA without being hampered by human contamination ( as P . vivax DNA is specifically amplified ) , provides a quantitative assessment of the proportion of each allele at the targeted SNPs , and can be easily customized to any parasite population or research question ( as different markers can be investigated by simply changing the PCR primers ) . We demonstrate the applicability of this approach by analyzing close to 500 P . vivax infections collected throughout Cambodia . Our analyses show that the data generated enable the characterization of genetic diversity and population stratification in Cambodian P . vivax , and the assessment of COI in this region , as well as its main determinants .
This study was conducted in accordance to the Institutional Review Board ( IRB ) National Ethics Committee for Health Research of Cambodia ( IRB 038NECHR ) and the Cleveland Clinic IRB ( IRB 12–936 ) . All patients provided written informed consent for the collection of blood samples , with the understanding that samples would be de-identified before processing . We collected blood samples from 401 P . vivax symptomatic patients from nine locations across Cambodia ( Fig 1 ) at three time points ( in 2004 , 2011 and 2013 ) and 88 asymptomatic patient samples from one district at two time points ( 2012 and 2013 ) . Note that different individuals were recruited at each time ( i . e . , no repeated sampling of the same individuals over time ) . Demographic and clinical data on the samples analyzed are provided in S1 Table . For each patient , we collected blood spots from finger prick or 5 mL of fresh blood in EDTA vacutainers and extracted DNA using a Qiagen Blood Mini Kit following the manufacturer’s protocol . We selected 128 SNPs among those for which both alleles were observed in at least two Cambodian infections that had been whole genome sequenced [15] . This condition was chosen to increase the chance that these SNPs were variable among Cambodian P . vivax . We used Primer3 v4 . 0 [16] to design primers targeting 100–300 bp around each selected SNP ( S2 Table ) . Each primer included a modified 5’ tail to allow barcoding and incorporation of the Illumina sequencing primers by a second PCR ( Fig 2 ) . We pooled 10 mM of primers from eight to ten primer pairs ( according to their melting temperature ) and generated a total of 13 primer pools . For symptomatic patient samples , we randomly assigned DNA into five 96-well plates and included duplicates for four samples as well as 25 water controls . For each plate , we conducted 13 individual PCR reactions ( i . e . , one primer pool per PCR reaction ) to amplify all 128 SNPs using for each PCR , 1 μl of DNA ( 30–50 ng DNA template ) in a 50 μl reaction consisting of 10 mM of each dNTP , 4 μl of 25 mM MgCl2 , 10 μl 5X GoTaq Flexi buffer , 10 μl of primer pool ( containing 10 μM of each primer ) , and 1 unit of GoTaq DNA Polymerase . Reactions consisted of an initial denaturation of 3 min at 94°C followed by 30 cycles of 45 sec at 94°C , 45 sec at 56°C , and 45 sec at 72°C , and a final 3 min extension at 72°C . PCR amplification products were visualized on 1% agarose mini-gels stained with ethidium bromide . We followed an identical procedure and randomly assigned asymptomatic patient samples into one 96-well plate containing three positive ( DNA from symptomatic samples ) and five negative ( water ) controls . For each plate we conducted 13 PCR reactions as described above to amplify all 128 SNPs , but using 45 cycles and Phusion High Fidelity Taq reagents . We combined the 13 PCR products from each individual sample and purified them using a QIAquick 96 PCR Purification Kit . We then performed a second PCR to add to each PCR product i ) Illumina adapters and ii ) a 6-nucleotide sequence specific to each individual sample ( Fig 2 ) . PCR conditions were identical to those reported above except for the number of cycles ( only 10 cycles were performed in this second PCR ) and the primer concentration ( 1 μl of each primer at 10 μM ) . We then pooled together the barcoded PCR products from 96 samples into a single tube and purified them using a Zymo DNA Clean and Concentrator kit . To remove primer dimers , we excised and purified bands between 200 and 600 bp from a 2% agarose gel using a Zymoclean Gel Extraction kit . Finally , we verified the library quality and quantity by Agilent Bioanalyzer-2100 and sequenced each library ( containing all PCR products from 96 samples ) on an Illumina MiSeq to generate 11–16 million 250 bp paired-end reads . We parsed all reads generated according to their respective 6-mer sequences and collapsed the paired-ends of each read into a single consensus sequence using PANDAseq [17] . We then mapped all consensus sequences to the P . vivax Salvador I strain reference genome sequence [18] using Bowtie2 v2 . 2 . 5 [19] . We eliminated from our analyses any targeted locus for which some of the amplified sequences mapped to more than one genome location , as this could indicate amplification of paralogous sequences . We also removed from further analyses loci that displayed more than two alleles at any given nucleotide position as this may also indicate amplification of paralogous sequences and , even if these alleles are genuine , would violate a common assumption of population genetic analyses ( the infinite site model ) . Finally , we discarded loci for which different alleles show systematic differences in coverage ( e . g . , if across all infections , haplotype A was always represented by more reads than haplotype B ) as this could indicate allelic dropout and poor amplification ( e . g . , the existence of polymorphisms at the primer sites ) . For symptomatic infections , we restricted our analyses to loci covered at ≥100 X in at least 100 samples . Individuals having less than 50 loci successfully genotyped based on the previous criteria were excluded . For asymptomatic infections , we included all individuals successfully genotyped ( ≥100 X ) at least 1 SNP . For each sample and each targeted SNP , we determined the coverage , and the frequencies of the reference and alternative alleles using samtools mpileup . Additionally , we determined the haplotypes at each locus for each sample by considering the combination of alleles present on the same read using the ‘MD’ string of the alignment files that summarizes all mismatch positions between the aligned reads and the reference genome . For both the targeted SNP genotypes and the locus haplotypes , we only considered alleles present in at least 10% of the reads of a given sample to avoid including PCR and sequencing errors . We classified samples based on their detected clonality: samples that carried a single allele at all positions ( i . e . , the minor allele at any given nucleotide was never observed in >10% of the reads ) were classified as monoclonal , while all other samples were considered polyclonal . For some analyses , we included , together with monoclonal infections , samples that displayed a single polymorphic site ( i . e . , only one nucleotide position showed two alleles ) as for these infections we can reliably determine the combination of unlinked alleles carried by one parasite at all positions ( i . e . , the “phased” genotypes ) . We referred to these infections as biclonal . We determined the combination of all major alleles across all initially targeted SNPs for the symptomatic infections ( i . e . , their “barcodes” [10] ) , to assess how informative these markers were at differentiating infections and calculated the mean number of nucleotide substitutions among them using MEGA v . 6 . 0 [20] . For these analyses , we only considered monoclonal and biclonal infections . We also determined the most likely number of clones present in each symptomatic infection using genotypes at the initially targeted SNPs and a maximum likelihood approach similar to the one implemented in COIL [21]: we estimated the population allele frequency at each initially targeted SNP using the mono- and biclonal infections and calculated , for each infection , the likelihood of the observed genotypes ( i . e . , either only the reference allele , only the alternative allele or both alleles at each SNP ) when the infection contained 1 to 5 unrelated parasites . We thus estimated the most likely number of clones present in each sample . Note that the reference allele frequencies are very similar between the mono- and biclonal infections on one hand and the polyclonal infections on the other hand ( S1 Fig ) confirming that all these parasites belong to the same population . We evaluated the correlations between different demographic and clinical parameters and COI using i ) the assessment of monoclonality based on the presence of more than one allele at any sequenced locus and ii ) the most likely number of clones in each infection . We tested the influence of geography ( western vs . eastern Cambodia , using the low transmission zone as the boundary ) , age ( <16 years of age vs . ≥16 years ) , sex ( male vs . female ) , sampling years ( 2004 , 2011 , and 2013 ) , glucose-6-phosphate dehydrogenase ( G6PD ) deficiency ( according to reported World Health Organization guidelines ) , hemoglobin ( Hb ) genotype , and hemoglobin concentration ( in g/dl ) [22] . We also evaluated the effect of parasitemia ( measured in parasites/μl ) on COI and whether parasitemia itself was influenced by the other demographic and clinical parameters . Finally , we tested whether the COI differed between symptomatic and asymptomatic infections . To account for the low genotyping success in asymptomatic samples , we randomly selected , for each asymptomatic sample , one symptomatic infection ( from the same location and year ) and used only the genotypes successfully typed in the asymptomatic sample to assess its clonality . We then compared the number of monoclonal infections in asymptomatic samples to the number observed in the randomly matched symptomatic samples . We repeated this procedure 1 , 000 times and calculated how often we observed more polyclonal infections in the symptomatic subsamples than in the asymptomatic infections . We assessed population stratification in Cambodian P . vivax with the program STRUCTURE v2 . 3 . 4 [23] and the entire haplotype information at the targeted loci using all symptomatic infections determined to be either mono- or biclonal . To increase the discriminating power of our analyses and facilitate the interpretation of the results , we included genotypes at the same loci inferred from whole genome sequencing data generated from parasites collected in Thailand ( N = 9 ) [24] and from non-Asian locations ( N = 11 ) [15 , 25–27] . We assigned the parasites into K = 1 to K = 4 populations using 100 , 000 burn-ins , 1 , 000 , 000 Markov chain Monte Carlo ( MCMC ) steps and a model of correlated allele frequency with admixture . For each K , we compared the results from five independent runs to verify convergence to a similar solution and recorded the proportions of ancestry of each parasite . We then tested whether the coefficients of ancestry of each parasite were significantly associated with geography ( western vs . eastern Cambodia and between countries ) , age ( child ( <16 years of age ) vs . adult ( ≥16 yrs . ) ) , and sex ( male vs . female ) using t-tests . An ANOVA was used to evaluate differences among sampling years . The sequence data are available in NCBI SRA under the BioProject PRJNA295043
We developed a multiplex genotyping assay ( Fig 2 ) that targeted 128 loci of 100–300 bp distributed throughout the P . vivax genome and covering all chromosomes ( Fig 3 ) . The loci were selected from a collection of SNPs found to be polymorphic among Cambodian P . vivax isolates that have been sequenced for their entire genome [15] . We applied this approach to analyze 489 P . vivax-infected individuals from Cambodia ( 401 symptomatic and 88 asymptomatic individuals ) . Out of the 128 primer pairs tested , one primer pair amplified DNA sequences that mapped to two different regions of the genome and was removed from the analyses . We also discarded one primer pair that amplified a low complexity locus causing misalignments in a homopolymer region . We additionally discarded six loci that displayed three alleles at some nucleotide positions ( possibly indicating the amplification of unannotated paralogous loci ) and 20 other loci that were not sequenced at 100 X in at least 100 symptomatic patient samples . The complete list of loci , their amplification primers and their inclusion in the final analysis is displayed in S2 Table . Overall , out of the 128 targeted loci , 100 were used in the final analysis . Four samples were analyzed in duplicate and showed very similar genotyping results ( S2 Fig ) suggesting that this approach did not generate many false positive calls . Of the 401 initial symptomatic patient samples , 99 samples were excluded from our analyses since we could only reliably genotype them ( i . e . , sequenced by >100 reads at each locus ) at less than 50 SNPs . Note that the genotyping success of symptomatic samples was not associated with parasitemia ( p = 0 . 23 ) , geographic origin ( p = 0 . 21 ) , or time of collection ( p = 0 . 38 ) . The samples included in the final analyses comprised infections from nine districts across Cambodia and three time points ( Fig 1 , S1 Table ) . The genotyping of asymptomatic patients was less successful and , out of the 88 initial samples , 43 were excluded from further analysis since they did not have 100 X read coverage at ≥1 SNP . P . vivax is haploid during the blood stage infection and monoclonal infections should therefore display a single allele at each nucleotide position . Using genotypes of the 100 SNPs that passed quality filters , 152 symptomatic patient samples , out of the 302 infections analyzed ( 50 . 3% ) , showed two alleles at , at least , one SNP , indicating the presence of two or more clones ( i . e . , polyclonal infections ) . For the remaining 150 samples ( 49 . 7% ) , the 100 SNPs all displayed a single allele and we were not able to rule out that these infections were monoclonal . Note that to limit the effect of amplification or sequencing errors , only alleles present in at least 10% of the reads were analyzed and therefore our analyses did not consider very minor clones . Since our genotyping assay relied on high-throughput sequencing of 100–300 bp surrounding each targeted SNP , it enabled discovery and genotyping of additional DNA polymorphisms located in proximity to the targeted SNP position . In the 100 loci successfully analyzed in the 302 symptomatic samples , we identified a total of 274 SNPs . Using all 274 SNPs , only 25 infections ( 8% ) were monoallelic at each nucleotide position ( i . e . , “monoclonal” ) while 277 ( 92% ) samples showed clear evidence of polyclonality . Of the latter , 60 infections ( 22% ) displayed two alleles at a single locus . We referred to these samples as “biclonal” and included them with monoclonal infections in a subset of our analyses since they enable inferring a “phased” multi-locus haplotype for these samples ( i . e . , determining which alleles are carried by the same clone ) . To assess how many genetic markers were necessary to rigorously identify COI , we randomly sub-sampled targeted SNPs and entire loci , and determined the proportion of infections deemed to be polyclonal using these data . Our results showed that single markers have limited power to identify COI and that more loci sequenced revealed more polyclonal infections , without apparent saturation ( Fig 4 ) . We also estimated the most likely number of clones present in each infection using a maximum likelihood approach similar to the one implemented in COIL [21] . On average , each Cambodian infection contained 1 . 81 clones with 1 . 65% of the infections displaying more than three clones . Note that this analysis relied only on the 100 initially targeted SNPs and therefore underestimated the true COI ( S3 Fig ) . We then investigated which parameters influenced COI in Cambodian P . vivax infections by i ) comparing monoclonal and polyclonal infections ( as determined by the genotypes at the 274 SNPs identified in the 100 loci sequenced ) , and ii ) comparing infections with different numbers of clones ( inferred statistically from the genotypes at the 100 unlinked SNPs ) . Using both data sets , we found that COI did not statistically differ between infections from western and eastern Cambodia ( Table 1 ) . Similarly , we found that COI was not significantly associated with parasitemia or the age or sex of the patient ( Table 1 ) . We also tested the association between various red blood cell phenotypes and COI but found no influence of hemoglobin level , hemoglobin genotype or G6PD deficiency ( Table 1 ) . Separating G6PD deficiency in mild and severe deficiencies , or restricting the analysis to male patients only ( that are haploid for G6PD ) did not change the results ( p>0 . 14 and p>0 . 49 , respectively ) . Note that , in this cohort , the parasitemia was not influenced by any of these parameters ( Table 1 ) . By contrast , we identified a significant association between COI and the time of collection ( p = 0 . 03 and p = 0 . 05 , respectively , for the comparison between monoclonal and polyclonal infections and the correlation between the year of collection and the most likely number of clones ) . The symptomatic infections collected in more recent years displayed significantly lower COI than older infections ( Fig 5 ) . The same pattern was observed by analyzing this association separately for each location ( S4 Fig ) . We then tested whether COI statistically differed between symptomatic and asymptomatic infections collected at the same location and the same time . To account for differences in genotyping success between symptomatic and asymptomatic samples , we randomly subsampled the symptomatic genotypes ( see Materials and Methods for details ) and showed that symptomatic infections were more often polyclonal than asymptomatic infections ( p = 0 . 04 ) . Population genetic analyses are complicated by the presence of multiple clones within an infection , which precludes reconstructing the genotype of each individual clone ( i . e . , the combination of alleles at different markers ) . We therefore limited our population genetic analyses to monoclonal ( N = 25 ) and biclonal ( N = 60 ) infections . Of these 85 parasites , 81 ( 95% ) had a unique combination of alleles at the 100 initially targeted SNPs ( i . e . , a unique “barcode” ) , with a mean difference of 19 . 6 nucleotides between two infections . The infections caused by genetically undistinguishable parasites were always observed in the same location: two patients from Veurn Say were infected by parasites with the same genotypes at all 100 SNPs , as were a pair of patients from Rovieng and two pairs of patients from Pailin . These apparently similar infections at the same location could be caused by the same parasite infecting different individuals or by highly related parasites ( e . g . , recombinant clones ) . Overall , these findings indicated that most infections were caused by genetically very different parasites suggesting a high level of genetic diversity in the Cambodian P . vivax population . We then assessed the population structure of Cambodian P . vivax using the entire DNA sequences of the 100 loci analyzed , treating them as multiallelic in the program STRUCTURE ( using only genotypes at the 100 unlinked SNPs initially targeted yielded , qualitatively , identical results but with a lesser resolution , R2 = 0 . 83 , p<2 . 2x10-16 ) . To improve the power of our analyses , as well as facilitate the interpretation of the results , we included genotypes at the same loci from P . vivax parasites that have been sequenced for their entire genome [15 , 24–27] . STRUCTURE assignments showed clear population separation between Southeast Asian ( Cambodian and Thai ) and South American and African parasites ( p<2 . 15x10-6; Fig 6 ) . Though Cambodian and Thai P . vivax showed ancestry from the same ancestral populations , they significantly differed in their coefficients of ancestry and appeared distinct ( p = 0 . 0001; Fig 6 ) . By contrast , we found no significant differences between parasites collected in western vs . eastern Cambodia ( p = 0 . 09 ) or among sites ( p = 0 . 09 ) , suggesting that the Cambodian P . vivax population is little stratified geographically . Note that the parasite population also appeared to be genetically diverse in this analysis: the genomes did not fall into a single homogenous population , but rather were split into multiple populations ( Fig 6 ) . Additionally , we observed no association between the parasite genotypes and the age ( p = 0 . 37 ) , sex of the patient ( p = 0 . 14 ) , or with the sampling year ( p = 0 . 79 ) ( Table 1 ) .
Successful elimination of P . vivax will require monitoring how the parasite population responds to control measures to identify areas where elimination is less efficient due to logistical or biological reasons ( e . g . , high vector density or emergence of drug-resistant parasites ) . Several studies have used assessment of genetic diversity to measure P . vivax demographic parameters ( e . g . , see [28–32] ) . However , due to the general paucity of genetic markers in P . vivax , most past studies have relied on a reduced number of microsatellite markers or SNPs , or on DNA sequences at highly variable surface protein antigens . Unfortunately , these markers only provide a partial perspective on the parasite population and could be biased by variations in mutation rates or the role of natural selection ( e . g . , positive selection on the locus investigated could , for example , be interpreted as evidence of population expansion ) . Here we described the development and implementation of a cost-efficient and high-throughput genotyping approach that is easily customizable and enables rigorous characterization of COI , genetic diversity and population stratification in the population of interest . The present assay targets over 100 SNPs distributed across the vivax genome , which minimizes potential artifacts in demographic inferences caused by the effect of natural selection driving the evolution of some markers . In addition , genotyping by high-throughput sequencing provides two additional key advantages . First , since 100–300 bp surrounding each targeted SNP are sequenced , the assay can identify additional DNA polymorphisms upstream or downstream of the targeted SNP position ( for a total of 274 SNPs in our study ) and provide haplotype data ( i . e . , multi-SNPs genotypes ) that are much more informative than single SNPs ( e . g . , see Fig 4 ) . Second , since more than 100 sequences are generated for each SNP and each sample , the genotyping is quantitative and enables high sensitivity and rigorous identification of alleles even if they are carried by only a minority of the clones ( a major difference compared to traditional genotyping approaches that have limited power to identify rare clones [33] ) . Note that , by simply changing the amplification primers used , the assay can be easily modified to target any locus of interest and could , for example , include the recently developed “P . vivax SNP barcode” [10] to enable direct comparison across studies . This flexibility is also important as it enables optimizing the experimental design to specific research questions . We chose here to analyze a large number of markers and stringent coverage cutoffs to rigorously investigate COI ( but with a relatively high rate of missing data ) . Alternatively one could reduce the number of markers or the depth of coverage to obtain a more complete dataset ( but with lower power to detect minor clones ) . Overall , we observed a very high proportion of complex infections among Cambodian patients: 92% of all symptomatic infections showed clear evidence of having multiple clones . In 60 infections , we only detected a single variable position . This observation could possibly be explained by the presence of two closely related clones in the infection , similar to what has been described in P . falciparum [34 , 35] . However , most of the remaining 217 infections ( 72% ) show numerous variable positions and likely represent instances of unrelated P . vivax clones simultaneous present in a given patient . This pervasive high COI contrasts with other studies that often show a large proportion of monoclonal infections ( e . g . , see [30–32] ) . While this discrepancy could be caused by lower parasite diversity or transmission rates compared to the situation in Cambodia , it is important to note that our estimates of COI are derived from sequencing numerous loci ( instead of a few microsatellites or SNPs ) , which dramatically increases our power to differentiate clones and therefore to detect complex infections ( Fig 4 ) . In particular , our findings are consistent with previous studies that used next-generation sequencing approaches [15 , 36–38] . It is interesting to contrast these patterns of COI in P . vivax with studies of P . falciparum: for example , a recent study , including more than 200 Cambodian P . falciparum collected in the same areas as our P . vivax infections , showed that these samples were essentially clonal [39] . In this regard , our study confirms previous findings based on a small number of microsatellites that showed striking differences in COI between sympatric P . falciparum and P . vivax infections collected in western Cambodia [40] . This difference is COI could reflect the biological differences between these species and , in particular , the role of the hypnozoites that might contribute to the elevated COI in P . vivax by releasing previously dormant parasites in the blood stream upon a new infection [41] . Alternatively , this difference could be caused by a recent decrease in P . falciparum diversity due to extreme drug pressure in this region ( though both parasite species are present in the same areas and are often treated indiscriminately ) . We made use of the large number of samples investigated ( N = 489 ) to test the influence of several patient and demographic parameters on COI . None of the patient data ( age , sex , hemoglobin level or genotype , G6PD deficiency or parasitemia ) were significantly associated with COI . In particular , higher levels of parasites in the blood were not statistically associated with polyclonal infections . Interestingly , we did detect modest but significant differences in COI between symptomatic and asymptomatic infections , with slightly more clones present in symptomatic infections . This observation will need to be confirmed by future studies but is consistent with the results of evolutionary models [42 , 43] . One could speculate that in P . vivax infections , which typically have a low parasitemia and multiple clones per infection , a higher number of clones could complicate the task of the immune system by presenting a more diverse range of surface antigens and lead to a more severe infection . In P . falciparum , increased COI has been associated with both lower and higher risk of malaria symptoms [44–47] . As more studies in both species become available , one will be able to rigorously address this interesting and clinically-important topic and test whether COI is associated with malaria symptoms and if the direction of this association is identical across Plasmodium species or in contrary , differs based on the specificities of each parasite ( e . g . , the differences in parasitemia and frequency of COI ) . Some of the patient parameters investigated in our study had been previously reported to influence P . vivax infections [48–51] . For example , some hemoglobin variants [52 , 53] and G6PD genotypes [54 , 55] might confer some degree of protection against P . vivax infections . By contrast , our analyses showed that these parameters do not seem to influence the number of clones infecting a given patient . This apparent discrepancy between the influence of these parameters on P . vivax infection and on COI could indicate that the infection complexity does not arise from multiple independent infections but , instead , from either simultaneous infection by multiple clones ( e . g . , the mosquito might directly release multiple parasites in the blood ) or by the relapse of dormant parasites from previous infections ( see e . g . , [56] ) . Interestingly , the geographic origin of the patients did not correlate with COI despite the very different incidence of P . vivax in western and eastern provinces: the vast majority of reported P . vivax cases ( 80% ) occurred in eastern Cambodia . By contrast , the year of collection was statistically associated with infection type: we observed that the proportion of monoclonal infections significantly increased in Cambodia from 2004 to 2013 . This increase in monoclonal infection may reflect a general decline in the Cambodian P . vivax population , which may be indicative of the success of elimination programs [57] ( e . g . , see [58 , 59] for a similar pattern in P . falciparum in Senegal ) . Indeed , during this period , the number of malaria cases in Cambodia has been halved , dropping from 113 , 855 cases in 2004 down to 56 , 271 cases in 2014 ( though a large component of this decrease can be attributed to a dramatic reduction in P . falciparum infections ) . This observation is encouraging for malaria control in Cambodia but analysis of additional samples collected more recently will be necessary to confirm this observation and to determine if this decrease is sustained over time . We observed a high level of genetic diversity in Cambodian P . vivax as illustrated by the large numbers of nucleotide differences differentiating clones , with , for example , an average of 19 . 6 differences , out of the 100 targeted SNPs , between two clones ( and up to 40 differences ) . This high diversity is also illustrated by the results of our population structure analyses: most parasite genomes do not cluster homogeneously into one population , but instead display complex patterns of ancestry from multiple populations ( Fig 6 ) . Interestingly , this extensive genetic diversity seemed to be distributed throughout Cambodia without clear demarcations: structure analysis failed to reveal any evidence of population stratification within Cambodia ( Fig 6 ) despite the existence of a low transmission zone separating our western and eastern sampling locations ( Fig 1 ) . Similar results have been reported for P . falciparum populations [60] . While the 100 loci sequenced provide sufficient information to rigorously differentiate parasites from Southeast Asia and the rest of the world , and to detect significant differences between Cambodian P . vivax and ( western ) Thai P . vivax , it is possible that more genetic markers ( or whole genome sequence data ) would be necessary to identify subtle population differences within Cambodia . Regardless , two possible demographic scenarios could explain the lack of strong population stratification . First , it is possible that the current Cambodian P . vivax population derive from a single diverse ancestral population that was only recently split into multiple separated populations , and that these have yet to differentiate significantly . Alternatively , our observation could be consistent with a high level of gene flow between different Cambodian P . vivax populations , resulting in high level of admixture ( which could explain the multiple ancestry patterns shown on Fig 6 ) . However , differentiating these hypotheses would require extensive whole genome sequencing data that would enable rigorously quantifying gene flow , population stratification and the demographic histories of the parasite populations . Despite progress in malaria control in general , vivax malaria remains an important public health concern in many endemic countries . The lack of tools to assess the efficacy of on-going control efforts hampers the rapid implementation of alternative or complementary approaches in the field . We described here a novel genotyping assay that can interrogate over 100 SNPs across the P . vivax genome , in a rapid , high-throughput and cost-efficient manner and can be tailored to specific polymorphisms or loci of interest . Using this approach , we were able to characterize geographical and temporal variations in P . vivax genetic diversity in Cambodia . Our analyses revealed that the proportion of monoclonal infections is increasing in Cambodia , which may be an indication of the success of the on-going control measures . However , our analyses also suggest that gene flow among Cambodian P . vivax is important , which could be problematic if antimalarial drug resistant alleles arise in one site as they could then inevitably spread throughout the region . Overall , this high-throughput genotyping technique adds new options to our existing toolkit for malaria control by providing a rigorous framework to assess how parasite populations respond to disease control strategies . | Plasmodium vivax is responsible for most malaria cases outside Africa but remains poorly understood . Here we describe a high-throughput assay that enables genotyping more than 100 SNPs in 96 samples simultaneously . We applied this assay to characterize P . vivax genetic diversity in ~500 individual infections collected throughout Cambodia and including multiple time points . Our analyses revealed that the Cambodian P . vivax population is very diverse genetically but geographically homogenous , without any obvious boundaries . We also observed that almost all infected individuals carried more than one P . vivax parasite , though the proportion of single parasite infections tends to increase in recent years . Our study illustrates how genetic information can be generated to monitor temporal and geographical variations in parasite populations and help assess the efficacy of malaria control programs . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [
"parasite",
"groups",
"medicine",
"and",
"health",
"sciences",
"plasmodium",
"population",
"genetics",
"geographical",
"locations",
"tropical",
"diseases",
"cloning",
"parasitic",
"diseases",
"parasitology",
"apicomplexa",
"molecular",
"biology",
"techniques",
"population",
"biology",
"genotyping",
"research",
"and",
"analysis",
"methods",
"artificial",
"gene",
"amplification",
"and",
"extension",
"molecular",
"biology",
"people",
"and",
"places",
"asia",
"genetics",
"biology",
"and",
"life",
"sciences",
"cambodia",
"malaria",
"evolutionary",
"biology",
"polymerase",
"chain",
"reaction"
] | 2016 | Complexity of Infection and Genetic Diversity in Cambodian Plasmodium vivax |
Ebola virus ( EBOV ) causes a severe hemorrhagic fever in humans and non-human primates . While no licensed therapeutics are available , recently there has been tremendous progress in developing antivirals . Targeting the ribonucleoprotein complex ( RNP ) proteins , which facilitate genome replication and transcription , and particularly the polymerase L , is a promising antiviral approach since these processes are essential for the virus life cycle . However , until now little is known about L in terms of its structure and function , and in particular the catalytic center of the RNA-dependent RNA polymerase ( RdRp ) of L , which is one of the most promising molecular targets , has never been experimentally characterized . Using multiple sequence alignments with other negative sense single-stranded RNA viruses we identified the putative catalytic center of the EBOV RdRp . An L protein with mutations in this center was then generated and characterized using various life cycle modelling systems . These systems are based on minigenomes , i . e . miniature versions of the viral genome , in which the viral genes are exchanged against a reporter gene . When such minigenomes are coexpressed with RNP proteins in mammalian cells , the RNP proteins recognize them as authentic templates for replication and transcription , resulting in reporter activity reflecting these processes . Replication-competent minigenome systems indicated that our L catalytic domain mutant was impaired in genome replication and/or transcription , and by using replication-deficient minigenome systems , as well as a novel RT-qPCR-based genome replication assay , we showed that it indeed no longer supported either of these processes . However , it still showed similar expression to wild-type L , and retained its ability to be incorporated into inclusion bodies , which are the sites of EBOV genome replication . We have experimentally defined the catalytic center of the EBOV RdRp , and thus a promising antiviral target regulating an essential aspect of the EBOV life cycle .
Ebola virus ( EBOV ) is a member of the genus Ebolavirus in the family of Filoviridae , and the causative agent of a severe hemorrhagic fever called Ebola virus disease ( EVD ) with case fatality rates of up to 90% [1] . While outbreaks are usually of comparatively small scale , the recent EVD epidemic in West Africa involved more than 28 , 000 cases with more than 11 , 000 deaths [2] , highlighting the urgent need for effective countermeasures against this virus . Significant progress has been made in recent years on the development of such countermeasures , with experimental vaccines showing promise in phase III clinical trials [3] . Similarly , a number of experimental therapeutics are under development , many of which target the viral polymerase L ( reviewed in [4 , 5] ) . This viral protein acts in concert with the other viral ribonucleoprotein complex ( RNP ) proteins , the nucleoprotein NP , the polymerase cofactor VP35 , and the transcriptional activator VP30 , to facilitate replication of the negative sense RNA genome of EBOV , as well as its transcription into viral mRNAs [6] . Despite its central role in the virus life cycle , relatively little is known about the L protein both in terms of its structure and in terms of functional details , which might in part be due to its large size and the fact that no specific antibodies are available , making biochemical studies of this protein challenging . Thus , much of what we know about L has been elucidated using reverse genetics-based life cycle modelling systems [7] . The most basic of these systems is the minigenome system [6] . Here a miniature version of the viral genome ( a so-called minigenome ) , in which all viral open reading frames have been removed and replaced by a reporter gene , but in which the non-coding terminal leader and trailer regions are retained , is expressed together with the RNP proteins in mammalian cells . These RNP proteins recognize the minigenome as an authentic viral template based on its leader and trailer regions , and replicate and transcribe it resulting in reporter activity levels that mirror these steps in the viral life cycle . As a modification of this classical monocistronic minigenome system , a so-called replication-deficient minigenome system has also recently been developed , which utilizes a minigenome with a deletion in the antigenomic replication promotor [8] . This system allows genome transcription to be investigated in isolation , something that is not possible in the classical system , where reporter activity is always the product of both transcription and genome replication ( which amplifies the number of genomic templates available for transcription ) . Further , with the tetracistronic transcription and replication-competent virus-like particle ( trVLP ) system another life-cycle modelling system has been developed . This system utilizes a minigenome that encodes not only a reporter , but also the viral proteins VP40 , GP1 , 2 , and VP24 , which are responsible for virus particle morphogenesis and budding , entry and fusion , and proper nucleocapsid assembly , respectively [9 , 10] . In this system minigenome replication and transcription in so-called producer ( p0 ) cells not only leads to reporter activity , but also to the formation of trVLPs , which package minigenomes-containing nucleocapsid-like structures and can infect target ( p1 ) cells . Using this system it is possible to model virtually the complete virus life cycle outside of a high containment laboratory . In all such minigenome-based systems an important consideration is to control for effects on plasmid-driven gene expression . This is usually done by including a control plasmid that encodes for another reporter , e . g . Firefly luciferase . This allows these cell-based assays to be normalized for well-to-well differences in transfection efficacy , cell density , or experimental effects , e . g . differences in cytotoxicity of tested drugs . Similarly , the extent to which virus RNP complex-specific genome replication and transcription are occurring is typically assessed by omitting the polymerase L . In the past , minigenome systems have been used to demonstrate the role of L in replication and transcription [6] , as well as to investigate the functional interactions of L with VP35 [11] . Indeed , this interaction has been shown to be crucial for genome replication and/or transcription , and a further study revealed an additional interaction of L with VP30 [12] . In that same study , we also identified a flexible linker region in L that is tolerant to insertions , and by subsequently fusing mCherry into this region we were able to characterize the intracellular fate of this protein , and show that it localizes in so-called inclusion bodies , which are formed in virus-infected cells and act as sites of genome replication [13] . However , fundamental aspects of L have still not been investigated , and for example the catalytic center of the RNA-dependent RNA-polymerase ( RdRp ) has not been experimentally determined , even though this information would be of great importance for drug development efforts [5] . In contrast , for other negative sense RNA viruses this catalytic center is well defined and involves a GDNQ motif , which based on bioinformatics analysis has been proposed to also be present and functional in the filovirus polymerase [14 , 15] . This motif represents a variant of the GDD motif found in RdRps of other viruses [16] , and sits in a deep channel of the polymerase [17] , where it complexes two metal ions essential for polymerase function [18] . Here , we provide the first experimental evidence that this motif is indeed essential for both virus genome replication and transcription , providing further insight into the molecular biology of filoviruses , and defining an important molecular target for the development of antiviral compounds against this deadly virus .
Filovirus reference sequences ( NC_014373 , NC_016144 , NC_001608 , NC_004161 , NC_006432 , NC_014372 , NC_002549 ) [19] as well as reference sequences for RSV ( NC_001803 ) and VSV ( NC_001560 ) were obtained from GenBank . Sequences were imported into Geneious v10 . 0 . 9 ( Biomatters ) , the L open reading frames were translated , and a multiple sequence alignment was performed using the ClustalW algorithm and the BLOSUM substitution matrix series . For calculating similarities in the pairwise distance analysis , amino acids were deemed similar if exchanges between those amino acids reached or exceeded a threshold of 0 in a BLOSUM62 matrix . HEK 293T ( human embryonic kidney; Collection of Cell Lines in Veterinary Medicine CCLV-RIE 1018 ) cells were maintained in Dulbecco´s modified Eagle’s minimum essential medium ( DMEM; ThermoFisher Scientific ) supplemented with 10% fetal bovine serum ( FBS; Biochrom ) , and 100 U/ml penicillin and 100 μg/ml streptomycin ( P/S; ThermoFisher Scientific ) . Huh7 cells ( human hepatoma cells; Collection of Cell Lines in Veterinary Medicine CCLV-RIE 1079 ) were cultured in a 1:1 mix of Ham’s nutrient mixture F-12 ( ThermoFisher Scientific ) and Iscove's modified Dulbecco's medium ( IMDM; ThermoFisher Scientific ) supplemented with 10% FBS , 100 U/ml penicillin and 100 μg/ml streptomycin . All cells were grown at 37°C with 5% CO2 . pCAGGS expression plasmids for NP , VP35 , VP30 , L , L-mCherry , T7 , firefly luciferase , Tim1 , and replication-competent and -deficient monocistronic minigenomes as well as the tetracistronic minigenome have been previously described [8 , 13] . A GFP-expressing minigenome was generated by deleting the luciferase open reading frame ( ORF ) from a luciferase-expressing minigenome and replacing it with the eGFP ORF using conventional PCR techniques and cloning with type IIS restriction enzymes . Primers and details of the cloning strategy are available upon request . Mutation of the EBOV-L gene ( specifically A13805C , A13807G , A13808C , A13811C , with all positions relative to the full length EBOV genome ) was performed by a combination of conventional and overlap extension PCR methods . To this end , first two separate touchdown PCRs using IProof polymerase ( Biorad ) and pCAGGS-L as template with the primers 5’-CCCGGGGCGGCCGCAAATG-3’ and 5’-CACCCATCACAGCTGAGCGTAACTTAAAAC as well as 5’-GTTTTAAGTTACGCTCAGCTGTGATGGGTGCCGCTGCGTGCATTACTGTTTTATC-3’ and 5’-GTTTGCCGAGTGTTAACTGTCCAAGG-3’ were performed . PCR-products were digested with DpnI ( New England Biolabs , NEB ) , and a second PCR was performed using the two PCR products as template , and the primers 5’-CCCGGGGCGGCCGCAAATG-3’ and 5’-GTTTGCCGAGTGTTAACTGTCCAAGG-3’ . The final PCR-product was cloned via NotI ( NEB ) and HpaI ( NEB ) into pCAGGS-L . To generate the mutant L fused to mCherry the region of pCAGGS-Lmut with the mutation was subcloned into pCAGGS-L-mCherry using the restriction enzymes HpaI and NotI . All plasmids were sequence confirmed by Sanger-sequencing . Minigenome assays were performed as previously described [8] , with slight modifications . HEK 293T cells were seeded into 12 well plates , and transfected at a confluency of about 50% using 3 μl Transit LT1 ( Mirus ) per μg DNA with expression plasmids encoding NP ( 62 . 5 ng ) , VP35 ( 62 . 5 ng ) , VP30 ( 37 . 5 ng ) , T7-polymerase ( 125 ng ) , firefly luciferase ( 12 . 5 ng ) , L or Lmut ( 500 ng ) or an equivalent amount of empty vector in the -L control , and a replication-competent minigenome ( transcription and replication assay ) or a replication-deficient monocistronic minigenome ( transcription assay ) ( 125 ng ) with Renilla luciferase as the reporter . At 24 hours post transfection ( p . t . ) , medium was exchanged against 2 ml of DMEM supplemented with 5% FBS and P/S , and after 48 hours p . t . luciferase activity was measured . To this end the supernatant was removed from the cells , 200 μl 1x Lysis Juice ( PJK ) was added to the cells , and after 10 minutes incubation at room temperature the lysate was removed and cell debris spun down 3 minutes at 10 , 000 x g . Then 40 μl clarified lysate was added to 40 μl Beetle Juice ( PJK ) or 40 μl Renilla Glo Juice ( PJK ) in black opaque 96-well plates , and luminescence was measured using an Infinite F200 PRO ( Tecan ) multimode reader with an integration time of 1 sec . Renilla luciferase activities were normalized to Firefly luciferase activities . To assess genome replication in isolation , a modified transcription and replication-competent virus-like particle ( trVLP ) assay [10] was combined with a newly developed RT-qPCR . HEK 293T producer cells ( p0 ) were transfected with expression plasmids for NP , VP35 , VP30 , the T7-polymerase , L , and a tetracistronic minigenome to generate trVLPs for infection of HEK 293T target cells ( p1 ) as previously described [10] . Target p1 cells in 12-well format were pre-transfected with pCAGGS-NP ( 62 . 5 ng ) , pCAGGS-VP35 ( 62 . 5 ng ) , pCAGGS-Tim1 ( 125 ng ) , as well as pCAGGS-L or pCAGGS-Lmut ( 500 ng ) . 24 hours p . t . these p1 cells were infected with 1 . 5 ml of clarified ( 5 minutes at 800 x g and room temperature ) , pooled p0 supernatant containing trVLPs . In order to do so , trVLP-containing supernatant was added to p1 cells , and the cells were centrifuged for 10 minutes at 1 , 000 x g , and then incubated at 37°C for 1 hour . After that , the inoculum was exchanged against 2 ml DMEM with 5% FBS and P/S . No VP30 was expressed in these p1 cells , so that only genome replication but not transcription could take place [6] . 48 hours post infection cells were lysed and RNA was isolated using the NucleoSpin RNA kit ( Macherey-Nagel ) following the manufacturer’s instructions . An additional DNA digestion step was performed using the turbo DNA-free kit ( ThermoFisher Scientific ) after RNA purification following the manufacturer’s instructions . RNA samples were then quantified by real-time RT-qPCR using the AgPath-ID One Step RT-PCR kit ( Applied Biosystems ) , with EBOV_IGR: 5’-6FAM-CAATAGCCAATACCAAACACCTCCTCCACAGCTTG-BHQ1-3’ as probe , and the primers EBOV_IGR-fwd2 5’-TCACAATCTACCTCTTGAAACAAGAAC-3’ and EBOV_IGR-rev2 5’-CATGACTTACTAATGATCTCTTAAAATATTAAG-3’ in 3 technical replicates , the results of which were averaged . To allow absolute quantification of copy numbers an RNA standard was prepared by in vitro transcription of the tetracistronic minigenome using the TranscriptAid T7 High Yield Transcription kit ( ThermoFisher Scientific ) following the manufacturer’s instructions , and quantified using a P-class P330 nanophotometer ( IMPLEN ) . 105 , 107 , and 109 RNA copies were used as standards in the real-time RT-qPCR . For western blot analysis Huh7 cells were seeded into 12 well plates and transfected as described above for the replication and transcription minigenome assay , with pCAGGS-L-mCherry , pCAGGS-Lmut-mCherry , or empty vector ( -L control ) in place of pCAGGS-L . After 24 hours p . t . the medium was changed to 1 ml medium supplemented with 5% FBS and P/S . The cells were lysed after 48 hours p . t . in 1x SDS sample buffer ( 10% glycerol , 5% 2-mercaptoethanol , 2% SDS , 37 . 5 mM Tris-HCl , 2 . 5 μg/ml bromphenol blue ) , incubated at 95°C for 5 minutes , and lysates were analyzed by SDS-PAGE and western blotting as previously described [20] using anti-mCherry ( Biozol: 1:1000 ) and anti-actin ( Sigma-Aldrich: 1:2000 ) primary antibodies and a peroxidase conjugated goat-anti-mouse secondary antibody ( Diavona: 1:10000 ) . For localization studies , Huh7 cells in 4-well μ-slides ( Ibidi ) were transfected with the same plasmids as for the western blot analysis but using half the amount of plasmid per well and a GFP-expressing minigenome . Additionally , 125 ng pmTurquoise2-H2A , which was a gift from Dorus Gadella ( Addgene plasmid # 36207 ) [21] , was cotransfected in order to label cell nuclei . Cells were visualized by spinning disc live cell microscopy using a Leica DMi8 with a Yokogawa CSU-W1 confocal scanning head , an Andor iXon Ultra 888 EMCCD camera , and 445 nm , 488 nm , and 561 nm laser lines . All images were taken using identical laser and camera settings for each wavelength . Paired two-tailed t-tests were performed using the GraphPad online QuickCalc ( https://www . graphpad . com ) .
For many negative sense RNA polymerases the catalytic center of their RdRp has been well defined , and incorporates a GDNQ motif [4] . Therefore , in order to identify the putative catalytic center of the EBOV polymerase , a multiple sequence alignment was performed between L open reading frames ( ORFs ) obtained from reference sequences for all filoviruses ( the ebolaviruses EBOV , Sudan virus ( SUDV ) , Bundibugyo virus ( BDBV ) , Reston virus ( RESTV ) , and Taï Forest virus ( TAVF ) , the marburgvirus Marburg virus ( MARV ) , and the cuevavirus Lloviu virus ( LLOV ) ) [19] , as well as Respiratory Syncytial Virus ( RSV ) and Vesicular Stomatitis Virus ( VSV ) . The alignment showed a very high degree of conservation among the filovirus polymerases ( Fig 1A ) , with the ebolavirus polymerases showing 90 to 94% sequence similarity and 73 to 84% sequence identity to each other , 79 to 80% similarity and 54 to 56% identity to LLOV L , and 70% similarity and 44% identity to MARV L . LLOV L showed a higher similarity and identity to the ebolaviruses polymerases ( 79 to 80% similarity and 54 to 56% identity ) than to MARV L ( 68% similarity and 43% identity ) , consistent with previous reports regarding the phylogenetic relationships between these genera [22] . As expected , the similarity to RSV and VSV was much lower , with 45 to 46% similarity and 15 to 16% identity between ebolaviruses and RSV , and 42 to 43% similarity and 13% identity between ebolaviruses and VSV . Nevertheless , a conserved GDNQ motif could easily be identified ( Fig 1B ) at positions 741–744 of the EBOV polymerase sequence ( position 815–818 in the multiple sequence alignment ) , and was identical in all analyzed sequences . After having identified a putative catalytic center within the EBOV L , we generated expression plasmids in which this motif was mutated by substitution of the DNQ sequence to 3 alanine residues . In order to assess whether these mutations affected expression of the protein , we performed western blot analysis after transient expression in 293T cells . Since no L-specific antibodies are available , we instead used L versions in which the fluorescent tag mCherry had been inserted into a flexible linker region , which we have previously shown tolerates insertions well without dramatically impacting protein function [13] . By western blotting we did not observe any significant differences ( p = 0 . 708 ) in expression level ( Fig 2A and 2B ) between L with an intact GDNQ motif and a mutated GAAA motif . In order to further assess whether there were changes to the intracellular localization of mutated L , we again used mCherry-tagged variants of L in combination with an eGFP-expressing minigenome . As expected based on previous studies , L-mCherry with an intact GDNQ motif localized into punctate structures that most likely represent early inclusion bodies ( Fig 2C ) [13] . Similar structures were also observed in the presence of L-mCherry with an abrogated putative catalytic domain . However , unlike the situation with untagged wild-type L , or L-mCherry with an intact GDNQ motif , we did not observe any GFP reporter activity with the L GAAA mutant , strongly suggesting a lack of activity of this mutant in transcription and/or replication . Given the strong impact of the GDNQ motif on reporter expression in context of the GFP-encoding minigenome , we next sought to quantify this impact using a Renilla luciferase-expressing minigenome , which allows easier quantification and more sensitive detection of reporter activity . To this end , we first performed classical minigenome assays , which measure both genome replication and transcription at the same time , but do not distinguish between these two steps . In this system , when using the L mutant with the abrogated GDNQ motif we observed a complete loss of reporter activity ( i . e . >1000 fold reduction ) with signals being reduced down to the background levels observed also in the complete absence of L , clearly indicating that this motif is absolutely essential for EBOV genome replication , transcription , or both of these processes ( Fig 3A ) ( -L vs . +L: p = 0 . 001; +L vs . +Lmut: p = 0 . 003 ) . When looking at the control Firefly luciferase , it became apparent that there was significantly less Firefly signal in the -L control than in the +L sample ( p = 0 . 019 ) . However , this difference was only about 2 . 4 fold , and did not contribute appreciably to the difference in the Renilla reporter signal , which was several orders of magnitude larger ( i . e . 1433 fold ) . In order to distinguish whether genome replication or transcription or both processes were impaired by mutation of the GDNQ motif , we next used a replication-deficient minigenome [8] . In these experiments absolute reporter levels were considerably lower , also in the positive control using wild-type L , compared to the reporter activity in the replication-competent minigenome ( Fig 3B ) , reflecting the strong contribution of minigenome replication to overall reporter activity in this system ( due to amplification of the vRNA templates available to serve as templates for transcription ) . However , again reporter activity in the presence of the mutated L was 67 fold lower than in the presence of wild-type L , which represents the background level for the assay ( -L vs . +L: p = 0 . 005; +L vs . +Lmut: p = 0 . 002 ) . This clearly indicates that the GDNQ mutation is essential for viral transcription , independent of any contribution from effects on viral replication . Again the Firefly signal appeared lower in the -L control compared to the +L sample; however , in this series of experiments this difference was not statistically significant . Finally , we also wanted to assess an independent impact of the GDNQ motif on genome replication . To this end , we developed a novel replication assay by combining a tetracistronic transcription and replication competent virus-like particle ( trVLP ) assay [10] with an RT-qPCR . To this end , a tetracistronic minigenome encoding VP40 , GP1 , 2 , and VP24 , in addition to a reporter , was expressed in p0 producer cells in the presence of the RNP proteins . This resulted in the formation of trVLPs that contain copies of the minigenome encapsidated in nucleocapsid-like structures . These trVLPs were then used to infect p1 target cells , which had been pretransfected with expression-plasmids for NP , VP35 , the EBOV adhesion factor Tim1 , and either wild-type L , or L with a mutated GDNQ motif . VP30 was intentionally omitted in p1 cells , since this protein has been shown to be an essential transcriptional activator , but not required for replication [6 , 8] . Total RNA from these p1 cells was harvested 2 days after infection , and subjected to an RT-qPCR assay targeting the intergenic ( i . e . non-transcribed ) region between the GP gene and the VP24 gene in the minigenome . Again , we saw a significant ( -L vs . +L: p = 0 . 048; +L vs . +Lmut: p = 0 . 011 ) reduction in vRNA/cRNA accumulation down to background levels when using L containing the mutated GDNQ motif , indicating that this motif is also required for genome replication .
The EBOV polymerase is the target for a number of potential antivirals such as favipiravir ( T705 ) [23] , BCX4430 [24] , GS-5734 [25] , and β-D-N4-hydroxycytidine [26] . Further , it has been the target for a number of high-throughput drug screens [27–29] , which generally exploit minigenome systems to allow rapid and easy modelling of the EBOV life cycle without the need for a high containment facility [30] . However , despite its central role in the virus life cycle , structural and functional data for this protein remains scarce . When the protein sequences of negative sense RNA virus polymerases of Rhabdo- and Paramyxoviruses were first published [31] , it quickly became clear that they share highly conserved regions that we now know to correspond to the RdRp , the polyribonucleotidyltransferase ( PRNTase ) , and the methyltransferase ( MTase ) domains ( reviewed in [32] ) , and for the Rhabdovirus VSV the structure of these domains has been solved at the atomic level [17] . The same conserved regions have since been tentatively identified based on sequence comparisons in other negative-sense RNA viruses including EBOV and MARV [33 , 34] . Further , while only limited crystal structure information is available for negative sense RNA virus polymerases , bioinformatics-based structural predictions suggest that the filovirus polymerase has a similar structure than polymerases from viruses for which a structure is known [14 , 15] . Experimental evidence of such a similar structure and experimental identification of molecular targets within the polymerase can help in rational drug design , as well as provide important insight in the mechanisms of action of compounds targeting L . This conservation of sequence and structure was the basis for the identification of flexible linker sites that allowed insertion of peptide tags as well as fluorescent proteins into the EBOV polymerase , with little impact on its expression , localization or function [12 , 13] , similar to previous studies involving the polymerases of paramyxoviruses [35 , 36] . Further , this assumption formed the basis for the multiple sequence alignment used in the present study to search for a putative catalytic center of the EBOV polymerase . Using this alignment , a GDNQ motif in the RdRp was readily identified , consistent with predictions by Cong et al . , who have suggested that D742 is a catalytic site in the filovirus polymerase [14] . After mutating the GDNQ motif , our functional results using luciferase-encoding minigenomes showed very clearly that this motif is required for genome replication and/or transcription of EBOV , and that this mutation completely abolishes transcription ( based on the results of the replication-deficient minigenome system ) and potentially both of these processes , resulting in reporter levels that are identical to samples completely lacking viral polymerase , and corresponding to the background noise of the luminometer ( about 102 RLU ) . Similar results were observed using GFP as a reporter , where in cells expressing L-mCherry with a mutated catalytic domain no GFP signal was observed . This was in contrast to cells expressing wild-type L-mCherry , where a strong GFP signal was readily observed , corresponding to robust minigenome transcription and replication ( albeit not in all cells , since in addition to L the other RNP proteins , as well as the minigenome , all have to be expressed in the same cell ) . In order to show definitively that genome replication is also abolished , we developed a replication assay by combining RT-qPCR technology and the recently published tetracistronic trVLP system . This approach has the advantage that neither minigenome-encoding plasmid DNA nor initial T7-transcribed and naked minigenome RNA is present in the p1 cells analyzed , since the source of the minigenome in those cells is infecting trVLPs which have packaged minigenome RNA-containing nucleocapsid like-structures [10] . Further , the target of the RT-qPCR is the VP30/VP24 intergenic region . In the tetracistronic minigenome this sequence is located between the GP1 , 2 ORF and the VP24 ORF ( in contrast , in the full-length EBOV the GP and VP24 genes are separated by the VP30 gene , so that no native GP/VP24 gene junction exists ) . This region harbors the longest non-transcribed sequence in the EBOV genome with a length of 144 nt [37] . This approach allowed us to exclude detection of mRNA , rather than cRNA/vRNA , despite the use of a one-step RT-PCR ( i . e . instead of a strand-specific two-step RT-PCR to target vRNA specifically ) . Additionally , we further exclude the erroneous detection of mRNA in this system by omitting expression of the transcriptional activator VP30 in p1 cells , as this protein has been shown to be required for transcription , but not for genome replication [6 , 8] . While , as with all point mutations , there is always the concern that the introduced mutations might negatively affect protein folding , we believe this not to be the case in this instance for two reasons: First , the mutated mCherry-tagged polymerase is readily recruited into inclusion bodies similar to those observed previously in cells infected with a recombinant EBOV expressing L-mCherry at early time points after infection , indicating that it still has to be able to interact with the other RNP proteins . Secondly , we have previously shown that an interaction with VP35 is required for stable expression of L , and that in the absence of VP35 L cannot be detected in significant amounts by western blotting [13] . The interaction domain between VP35 and L has been mapped to the amino acids 280 and 370 , which are located in the RdRp domain of L [11] . Since we do not see any differences in the expression level of our mutated L , we have to conclude that this mutant remains able to interact with VP35 , and that , therefore , the RdRp domain which harbors the mutation is not grossly misfolded . On a technical note , analysis of the Firefly control luciferase in this study showed signals for this reporter that were lower in the -L controls than in the +L samples ( regardless of whether L was functional or not ) . This phenomenon is most likely due to the fact that it is common good practice to include empty vector in samples where plasmids are omitted for experimental reasons ( e . g . in -L controls ) , in order to equalize the transfected plasmid mass . However , given the size of the L expression plasmid ( 11 . 6 kB ) vs . the empty vector ( 4 . 8 kB ) , this means that in terms of absolute numbers many more empty plasmids than L expression plasmids are transfected , which may lead to a reduction in gene expression from the other co-transfected plasmids . This effect can skew the results of minigenome assays , since reporter luciferase activity values are normalized to these control luciferase values , thus artificially inflating -L control values . While this effect is small compared to the very large dynamic range of EBOV minigenome assays ( which in our hands is about 3 log10 ) , particularly for high-throughput assays were a large dynamic range is required and this control luciferase is essential to normalize for well-to-well variations , this situation is less than ideal . In contrast , when using the mutated L version , no differences in plasmid-based gene expression ( i . e . Firefly controls ) are observed , while genome replication and transcription are completely abolished . Thus , this mutant represents a superior control compared to the -L control , particularly in context of high-throughput assays . This is of particular importance as high throughput-screens under BSL4 conditions , which are necessary for work with infectious EBOV , are significantly more complex and cost-intensive than similar screens under BSL2-conditions , providing a strong incentive for the use of EBOV minigenome and other life cycle modelling systems for drug screening purposes [30] . Further , the development and use of similar catalytically inactive polymerase mutants in the place of conventional -L controls may represent a technical improvement for other minigenome systems ( e . g . for other viruses ) that may demonstrate more modest dynamic ranges and thus be more significantly impacted by such effects . Overall , we have experimentally confirmed the catalytic center of the RdRp of the EBOV polymerase , which represents a promising target for the development of antivirals . This work provides a basis for future studies aimed at inhibiting the activity of this protein , which is absolutely crucial for the virus life cycle , as well as providing technical advancements in the tools available for high-throughput screening applications . | Ebola viruses cause severe hemorrhagic fevers , and were responsible for the devastating Ebola virus disease epidemic in West Africa from 2013 to 2016 . While a number of experimental therapeutics against these viruses target the viral polymerase , there are still significant gaps in our knowledge regarding this essential viral protein . In particular , until now no experimental evidence has been provided identifying the catalytic center of the viral RNA-dependent RNA polymerase , which is absolutely essential for the virus life cycle due to its role in replicating and transcribing the viral negative-sense RNA genome . Based on a comparison to related negative-sense RNA viruses from other virus families we identified a putative catalytic center within the Ebola virus polymerase , and provide the experimental evidence that the Ebola virus polymerase indeed utilizes a classical GDNQ motif for both genome replication and transcription . This finding not only increases our knowledge regarding the molecular biology of Ebola viruses , but also defines a molecular target for the development of antivirals against this deadly virus . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"luciferase",
"enzymes",
"dna-binding",
"proteins",
"microbiology",
"negative-sense",
"rna",
"viruses",
"enzymology",
"viruses",
"developmental",
"biology",
"polymerases",
"rna",
"viruses",
"sequence",
"motif",
"analysis",
"research",
"and",
"analysis",
"methods",
"sequence",
"analysis",
"sequence",
"alignment",
"bioinformatics",
"proteins",
"oxidoreductases",
"life",
"cycles",
"viral",
"replication",
"biochemistry",
"virology",
"database",
"and",
"informatics",
"methods",
"ribonucleoproteins",
"biology",
"and",
"life",
"sciences",
"organisms"
] | 2017 | Characterization of the catalytic center of the Ebola virus L polymerase |
Pluripotent embryonic stem cells ( ESCs ) have the unique ability to differentiate into cells from all germ lineages , making them a potentially robust cell source for regenerative medicine therapies , but difficulties in predicting and controlling ESC differentiation currently limit the development of therapies and applications from such cells . A common approach to induce the differentiation of ESCs in vitro is via the formation of multicellular aggregates known as embryoid bodies ( EBs ) , yet cell fate specification within EBs is generally considered an ill-defined and poorly controlled process . Thus , the objective of this study was to use rules-based cellular modeling to provide insight into which processes influence initial cell fate transitions in 3-dimensional microenvironments . Mouse embryonic stem cells ( D3 cell line ) were differentiated to examine the temporal and spatial patterns associated with loss of pluripotency as measured through Oct4 expression . Global properties of the multicellular aggregates were accurately recapitulated by a physics-based aggregation simulation when compared to experimentally measured physical parameters of EBs . Oct4 expression patterns were analyzed by confocal microscopy over time and compared to simulated trajectories of EB patterns . The simulations demonstrated that loss of Oct4 can be modeled as a binary process , and that associated patterns can be explained by a set of simple rules that combine baseline stochasticity with intercellular communication . Competing influences between Oct4+ and Oct4− neighbors result in the observed patterns of pluripotency loss within EBs , establishing the utility of rules-based modeling for hypothesis generation of underlying ESC differentiation processes . Importantly , the results indicate that the rules dominate the emergence of patterns independent of EB structure , size , or cell division . In combination with strategies to engineer cellular microenvironments , this type of modeling approach is a powerful tool to predict stem cell behavior under a number of culture conditions that emulate characteristics of 3D stem cell niches .
Pluripotent embryonic stem cells ( ESCs ) have the unique ability to differentiate into cells of the three germ lineages that form all of the tissues and organs of a mature organism . Differentiation of pluripotent ESCs can be induced in vitro via a variety of existing approaches to emulate aspects of the developmental program . One of the most widely used techniques relies upon the formation of multicellular aggregates composed of undifferentiated ESCs in suspension culture , commonly referred to as embryoid bodies ( EBs ) [1] , [2] , that spontaneously induce the differentiation of ESCs within the 3D aggregate [3] , [4] . Due to the fact that EBs mimic the physical structure and cellular composition of the morphogenic embryonic microenvironment , they have been used to study aspects of development in vitro as well as the formation of primitive tissue complexes [3]–[5] . Despite the utility of the approach , robust methods to control EB differentiation in vitro remain limited due to an incomplete understanding of the complex interactions within the 3D multicellular aggregates that mitigate cell fate decision [6] , [7] . The development of techniques to control ESC differentiation in vitro requires an improved understanding of cellular cues that regulate differentiation and the means to precisely control these complex signals . Considerable effort has focused on ascertaining the role of individual components of the cellular microenvironment in regulating cell fate decisions . The extent to which cell-cell communication via paracrine [8] , [9] , autocrine [9]–[11] , or direct contact signaling [12]–[14] enhance or inhibit differentiation have been investigated in various contexts . Exogenous manipulation has been used to control differentiation by the addition or removal of various soluble factors in a temporally regulated manner in an effort to mimic morphogenic cues . Factors that preserve pluripotency ( e . g . LIF [15]–[17] , BMP-4 [15] ) and factors that can initiate differentiation ( e . g . Activin A [18] , FGF-2 [18] , and retinoic acid [19] ) have been thoroughly examined , both in terms of the appropriate doses and their temporal administration . In many cases , the signaling pathways and modes of action of these growth factors are also known but the effects of combinatorial treatments can be difficult to predict a priori [16] , [20] . Although exogenous factors have proven necessary for the in vitro and in vivo maintenance or differentiation of ESC populations , they are not the only factors regulating stem cell behaviors . The biochemical composition of the cellular microenvironment [9] , [21] and extracellular matrix ( ECM ) [22]–[24] have also been implicated in the regulation of cellular niches . In addition , the mechanics and physical properties of the microenvironment can also impact cell phenotype [25] . Given that cell fate transitions occur in complex environments where biochemical and physical cues coexist , elucidating the role each of these combinatorial factors via experimental studies alone remains a significant challenge . Therefore , although the aforementioned studies can provide information about certain individual factors in isolation , new approaches that allow systematic investigation of combinations of parallel factors that regulate stem cell differentiation are needed to more accurately predict cell response to complex microenvironmental cues . In many instances , computational modeling strategies have been successfully used to recapitulate the integration of complex signals that direct cell fate decisions and correctly predict the resulting phenomena . Depending on the desired resolution of the system , ordinary differential equations can be used to model a variety of processes in stem cells including - but not limited to - cellular signaling events [26]–[28] , protein interaction networks [29] , and genetic networks [30] . Partial differential equations can be used to assess spatial changes introduced via diffusion of molecules; this approach has been extensively studied to examine gradients of nutrients in cancer cell spheroids [31] , as well as mass transfer limitations in EBs [32] . Alternatively , to model the structure of cellular aggregates [33]–[35] , cellular division and tissue formation [36]–[43] , and pattern formation in biological systems [38] , [39] , agent-based modeling has been applied to overlay rules-based and physical modeling approaches [34] . Moreover , agent-based models have been used to investigate dynamic processes of multicellular systems , such as morphogenesis [43] , [44] and formation of physical tissues [45] . Investigation of the spatial and temporal regulation of stem cell differentiation using agent based model approaches has not been attempted , yet the ability to examine how structural features of the stem cell niche influence the spatial patterns associated with loss of pluripotency is attractive for studying differentiation in 3D EB systems . This study demonstrates the utility of computational rules-based modeling to predict emergent spatial patterns associated with one pluripotent transcription factor in EBs ( Oct4 ) and investigate macroscopic principles that can play important roles in determining cell fate transitions .
Embryoid bodies ( EBs ) are 3D multicellular spheroidal aggregates that self-assemble via E-cadherin mediated interactions in suspension conditions [46] , [47] . Our first goal in constructing a model description of EBs was to accurately recapitulate the overall multicellular structure based on the physical properties of individual mouse embryonic stem cells ( mESCs ) . Prior models of multicellular structures have described the individual cellular agents as incompressible objects consisting of ellipsoids [48] . We opted to use a physics-based modeling approach in which cells were modeled as incompressible rigid spheres as this is a powerful and portable method for representing complex aggregate shapes . To determine if modeling mESCs as spheres was appropriate , the effective surface area ( Fig . 1A ) and radii ( Fig . 1B ) of individual ESCs were experimentally determined via Coulter counter analysis . The average surface area to volume ratio of the mESC line was 3 . 26+/−0 . 15 , which is only ∼8% higher than the theoretical value of 3 . 00 for a spheroid . Due to the increased computational costs associated with an ellipsoid collision detection algorithm and the relatively low error in the surface area-to-volume ratio ( <10% ) , we proceeded by representing each cellular agent as a sphere . The distribution of cell radii from the Coulter counter measurements ( Fig . 1A , B ) were used to create the population of spheres for each agent in the EB simulations . These cell agents were randomly seeded into a box , which served as an initial boundary for the simulation , and then forced to aggregate using a gravitational point source into a multicellular spheroidal structure . The structures of in silico and in vitro aggregates were assessed for aggregates of 250 and 1000 cells using four parameters: radius , circularity , connection count , and connection lengths . Size and circularity were used to assess the entire aggregate structure and were experimentally determined through the analysis of phase contrast images , while similar measurements were obtained using projections of the in silico EBs onto a 2-dimnesional plane ( Fig . 1D ) . The results indicated that the model appropriately captured the macroscopic features of the relative EB aggregates since there were no statistical differences between the model and experimental metrics ( Fig . 1E , 1F ) . The connection count and connection length parameters were calculated from the spatial distribution of individual cells comprising the aggregates and serve as quantifiable metrics for assessing local micro-scale organization within EBs . These parameters were assessed by individual cell labeling performed in confocal microscopy images and via computational algorithms for the in silico EBs ( Fig . 1G ) . As an example , the blue box in Fig . 1G highlights a cell with an average connection length of 14 . 87+/−2 . 07 microns and connection number of 4 . For in silico EBs , the aggregates were “virtually sectioned” ( at a 10 µm thickness ) to perform similar analysis on a 2D projection , and neither the average circularity nor the connection lengths differed statistically from the experimentally derived EB values ( Fig . 1H , 1I ) . Overall , these results quantitatively comparing four different physical parameters indicated that the model was able to accurately create the structure of individual EBs on both the aggregate and cellular scales , providing an accurate structural framework for our subsequent analysis of spatial patterning . Throughout the subsequent discussion of the results , pluripotent cells that exhibit loss of Oct4 expression are simply referred to as “differentiated” , acknowledging the caveat that Oct4− cells are not terminally differentiated . As Oct4 is concomitant with loss of pluripotency , it was used to monitor the pluripotent state of the cells [49]–[52] . This process has been modeled as a bi-stable transition , which causes an all or none response [52] . The temporal patterns of loss of pluripotency were evaluated in 250- and 1000-cell EBs via confocal microscopy to examine Oct4 expression . Starting from a homogeneous population of undifferentiated cells , spatial heterogeneity ( as defined by loss of Oct4 expression ) was observed over the course of the multiple days of evaluation . In order to capture the diversity of spatial pattern heterogeneity , a classification system was developed . Based on preliminary results from both the experimentally derived EBs and the model , six different categories of patterns were proposed: Oct4+ , inside-out , outside-in , connected , random , and Oct4− ( Fig . 2 ) . These patterns can be loosely grouped into three larger categories: Oct4+ , transitioning , and Oct4− . The outside-in , inside-out , connected , and random patterns were all considered transition patterns as they captured intermediate stages of the differentiation process . Inside-out patterns are characterized by differentiation in the middle of the EB and undifferentiated cells on the outside; conversely outside-in patterns exhibit differentiation on the outside , and undifferentiated cells in the middle . Connected patterns were defined as multiple distinct connected regions of cells of the same state , whereas random patterns were classified as no identifiable pattern based on a lack of connectivity . In the smaller 250-cell EBs , Oct4 expression persisted for up to six days ( Fig . 3A ) . Rapid loss of Oct4 was observed between days 3 and 5 ( Fig . 3C ) and the patterns associated with differentiation were classified entirely as “connected” ( Fig . 3D ) . In 1000-cell EBs , differentiation patterns were assessed over a 7-day period ( Fig . 3B ) . Differentiation was observed to occur at a later time than the smaller 250-cell EBs , with transition patterns occurring from days 4 to 7 ( Fig . 3E ) . The spatial patterns in the 1000-cell EBs associated with differentiation were more varied than the 250-cell EBS but also were primarily classified as “connected” ( Fig . 3F ) . At each time point , pattern classification for each EB size was performed to generate temporal differentiation profiles for each time point ( Fig . 3D , 3F ) . The trajectories of differentiation were calculated by assessing how the number of differentiated , undifferentiated , and transitioning patterns changed over time . Although the types of patterns associated with differentiation only changed slightly with EB size ( Fig . 3D , 3F ) , the kinetics of the process did appear to change appreciably . The 250-cell EBs began differentiating at ∼ day 3 and finished within one day , whereas the 1000-cell EBs started a similar process later at ∼ day 4 , and took nearly 3 days to fully exhaust Oct4 expression ( Fig . 3C , 3E ) . After validating the generation of an appropriate 3D geometry of EB aggregates , rules-based modeling was performed by creating network structures , in which “nodes” represented individual cells , and “connections” represented physical interactions between adjacent cells; nodes were allowed to convey information with the macrostructure along the defined connections . The goal was to determine if simple rules accurately produced the distribution of spatial patterns observed experimentally . During these initial simulations the macro-structures were assumed to be static ( i . e . no proliferation , migration or apoptosis ) . Cells could exist in either of two states: undifferentiated ( Oct4+ ) or differentiated ( Oct4− ) . The transition between these two states was chosen as binary based on previous modeling work [52] and occurred based on different rule formulations: “random” , “positive feedback” or “competing feedback” ( Fig . 4 ) . The random rule configuration represented a stochastic , basal differentiation probability ( Fig . 4A ) . The positive feedback rule was based on a paradigm in which differentiated cells bias neighboring cells to differentiate ( Fig . 4B ) and was inspired by differentiation induced via direct cell-cell interactions [13] , [53] . Finally , the competing feedback rule depicts a situation where differentiated cells promote subsequent differentiation of neighboring cells while undifferentiated cells inhibit this transition ( Fig . 4C ) . Positive feedback in this rule was based upon the known role of soluble factors to maintain pluripotency [54] , while negative feedback comes from the differentiation induced via the cell-cell interactions discussed above [13] , [53] . Representative outcomes for each of the rules are shown for 250 ( Fig . S1 ) and 1000 cells ( Fig . 5 ) per EB . The patterns represented by different rules did not differ significantly across EB sizes ( Fig . 5 and Fig . S2 ) ; however , differences in the distribution of the patterns were observed between different rule configurations . The random rule transitioned through largely random patterns ( Fig . 5I , S2I ) , whereas both the positive feedback and the competing feedback exhibited an enrichment in the connected patterns ( Fig . 5F , S2F and Fig . 5C , S2C respectively ) . Because the experimental EBs differentiated largely through connected patterns ( Fig . 3D , F ) based on the pattern classification , it was difficult to evaluate which rule configuration ( s ) best emulated the experimental data . To ensure these results were not an artifact of the multicellular aggregate structure , resultant patterns of 250 and 1000 cell EBs were evaluated for each of the three rules ( Fig . S2 ) across numerous different simulated structures without discernible changes in outcome . To glean insight into the evolution of the connected patterns , we used two quantitative metrics , undifferentiated cluster number ( UCN ) and differentiated cluster number ( DCN ) , to assess pattern formation and simulation trajectories against a normalized time ( τ ) axis . Analyzing the cell phenotype transitions by the UCN and DCN metrics revealed distinct paths of pattern formation for each of the different rules . The qualitative shape of the trajectories was independent of EB sizes ( Fig . S3 ) . From such curves , critical points ( τ = . 2 , τ = . 4 , and τ = . 6 ) representing rapid changes or important regions across all rules were chosen and representative EB slices were displayed ( Fig . 5B , E , H ) ) . Analysis of the trajectories themselves revealed insight about the types of clusters being formed in the “connected” patterns . In the “positive feedback” scenario , the loss of Oct4 expression was characterized by a high number of differentiated or undifferentiated clusters , suggesting localized intercellular neighbor influences regulating phenotype transition ( Fig . 5D , Fig . S1D ) . The peak in UCN at τ = . 6 was characterized by a large number of isolated pockets of undifferentiated cells . In contrast , the “competing feedback” rule peaked through a high number of differentiated clusters , but never amassed a high undifferentiated cluster number ( Fig . 5G , Fig . S1G ) which matched the larger isolated and persistent clusters of Oct4 positive cells experimentally observed in both the 250- and 1000-cell EBs . Taken together , these data indicate that the “competing feedback” rule matched the patterns observed biologically with the highest fidelity for the different size EBs examined . Furthermore , the trajectory analysis provided novel information about the evolution of certain patterns . For example , random differentiation can be characterized by a high initial spike in the DCN as this signifies the emergence of several small clusters of differentiated cells ( Fig . 5A , Fig . S1A ) , and the duration of this spike represents how long the random patterns persist throughout the duration of the model . If the UCN remains fairly low , the pattern transitions into a connected phenotype , again evidenced by the low number of undifferentiated clusters of cells ( Fig . 5B , Fig . S1B ) . When the UCN remains at 1 , this signifies either an inside-out , or outside-in pattern . However , if the UCN transitions towards a high value , this signifies that the differentiation is still largely governed by random patterns ( Fig . 5A , Fig . S1A ) . Overall , these metrics provide quantitative metrics for assessing the types of patterns formed , and the evolution of these patterns over time . One limitation of the model , however , was that the kinetics of pattern formation could not explain the differences in kinetics experimentally observed between different EB sizes ( Fig . 3D , F ) . This suggested that although a static size aggregate modeling approach was sufficient for predicting the prevalence of different spatial pattern classifications , it did not fully capture the kinetics of experimental Oct4 loss . In order to further investigate the kinetics of the pattern transitions , we modified our modeling framework to include cell division and embryoid body growth . We hypothesized that dynamic processes , such as cell division and growth of the EB aggregate , influence spatial patterns of Oct4 expression loss . To investigate the effect of cell division on this loss of pluripotency transition , a revised model which could simulate growing EB structures was created in which the Oct4+ cells divided at a faster rate than the Oct4− cells . With this approach , cells were modeled as rigid spheres while cell-cell connections were represented by springs which helped maintain the overall macro-structure of the aggregate ( Fig . 6A ) . We observed experimentally that the size of differentiating ESCs did not change appreciably with time , which reduced complexity from the model description ( Fig . 6B ) . The first step was to determine an estimated cellular division rate for Oct4+ and Oct− cells . This was accomplished using experimental growth data approximated from the size of the embryoid bodies ( Fig . 6C ) , yielding a rate of ∼18 hours for division of our stem cells and ∼51 hours for division of the differentiated cells which is consistent with the literature [55] . Using these division rates , we applied the rules we derived in the former static model to the new dynamic model . As an internal control , we ran division simulations with no rules to monitor any bias introduced by the model ( Fig . S4 ) and found that the cells grew in an exponential manner ( Fig . S4A ) , while the density of the aggregates remained constant ( Fig . S4C ) . Furthermore , the average connection number and connection lengths of the network remained constant with time ( Fig . S4 E ) . Connection length remained constant as a function of the aggregate radius , whereas the connection number decreased on the outer layer of the aggregate , as was expected ( Fig . S4E ) . Taken together these results suggest that structurally no bias was introduced into the model by introducing cell division . Next we examined the spatial patterns formed under the various rule configurations during enlargement of the EB over a 5–7 day period , with the pattern trajectories now normalized by the cell number to account for cell growth . Trajectories simulated over the 6 day culture period indicated consistent pattern distributions and their evolution over time ( Fig . 6E–F ) . While few differences were observed in the trajectories as a function of EB size , the cumulative variance in the UCN and DCN metrics was greater for the 250 cell EBs ( Fig . 6G ) , suggesting a more heterogeneous population of differentiated cells . At later time points ( >Day 5 ) the positive feedback rule generated clusters of undifferentiated cells predominantly on the outer edges of the EB . In contrast , the competing feedback rule produced larger clusters of undifferentiated cells localized towards the center of the EB . The trajectories of these growth simulations matched the general shape of the trajectories for the static simulations ( Fig . S3 ) , with a notable exception of the random rule . Analysis of the percentage of undifferentiated trajectories revealed that over half the cells remained undifferentiated at the seven day time point , which may explain the absence of the DCN to UCN transition observed in the previous static model simulations . The motivation behind modeling a dynamic EB structure was to more closely recapitulate the emergent morphogenic processes occurring over the transitional 5–7 day period and to investigate whether the inclusion of cell division and EB enlargement influenced the rate of emergent spatial patterns ( Fig . 6H ) . In the case of the random and competing feedback rules , no observable difference in the percentage of differentiated cells appeared , however the positive feedback rule resulted in the percentage of undifferentiated cells being lower in the 1000 cell versus 250 cell EBs , likely due to the total number of cells present . Taken together these results suggest that differential cell division does not significantly influence the formation or evolution of phenotype patterns over time and the pattern formation process is dominated by the regulatory mechanisms encoded in the probabilistic rules .
Differentiation is a complex biological process involving the coordinated regulation of multiple genes by intrinsic and extrinsic factors . Rather than attempt to model the intricate network of genetic circuitry , signaling mechanisms , and environmental cues , we approached the loss of pluripotency from a simplified perspective designed to elucidate the most basic principles dictating pattern formation in a spherical multicellular system . We developed a modeling framework capable of recapitulating the physical properties of embryoid bodies for multiple sizes and under conditions of cell division and aggregate growth . This framework allowed us to simulate how different probabilistic rules were manifest in the emergence of spatial patterns and to examine the evolution of these patterns over time . Through comparison of the simulated pattern trajectories with the pattern classification developed for our experimental data , we determined that all possible pattern classes - as well as a similar distribution within these classes - are possible with the agent rules employed . Furthermore , from both static and dynamic frameworks , our simulations indicate that aggregate structure , size , and growth are physical features that do not dictate the distribution of spatial patterns or their trajectories as a function of time . Differentiation is classically thought of as a binary transition from the undifferentiated pluripotent stem cell state to differentiated phenotypes in embryoid body ( EB ) models [5] , [31] , [46] . Models of stem cell differentiation often consider these events on a population basis [56] or at an intracellular signaling level [57] . Here we have shown that the transition from Oct4+ to Oct4− states produces dynamic , spatially heterogeneous patterns in a continuous manner . These results indicate that modeling embryonic stem cell ( ESC ) fate decisions as a stochastic , binary process is sufficient to predict the dominant emergent spatial patterns of loss of pluripotency . Additionally , 250- and 1000-cell EBs undergo loss of pluripotency through the same intermediate patterns which suggests that the macro processes governing this early stem cell transition ( while occurring at slightly different rates ) are largely independent of size . In addition , comparing static structures and cell division models revealed that spatial transition patterns and their evolution are not significantly affected by cell growth . Our modeling approach created in silico aggregates with similar properties to in vitro EBs . The spring-based constraint representing cellular adhesions was able to accurately capture the evolution of the aggregate architecture . Additionally , the high-level rules described here were able to reproduce the emergence of a variety of spatial patterns , all of which could be observed experimentally in EBs . Both modeled and experimental EBs demonstrated enrichment in connected patterns of cells . Quantification of model simulation outputs by differentiated and undifferentiated cluster number ( DCN and UCN ) allowed visual representation of the time evolution of connected patterns in 3D multicellular aggregate systems . The use of the UCN and DCN metrics provided information about not only the types of connected patterns formed , but also the transitions between the different types of spatial patterns . A combination of analyses using these metrics and manual pattern identification indicated that the “competing feedback” rule scheme that accounted for opposing influences of neighboring pluripotent and differentiated cells produced a distribution of spatial patterns that most closely resembled the experimentally observed spatial patterns for both 250- and 1000-cell EBs . This was observed in both the static and dynamic division models , and could be hypothetically represented biologically as a combination of cell-cell signaling and the complex interplay of local soluble factors and other chemical gradients that influence pluripotent cell fate decisions . [17] , [53] , [58] , [59] . However , this is only one possible explanation for a coarse-grained description , and a variety of other signaling pathways and molecules are likely involved in regulating this transition . Interestingly , none of the rules were able to explain the observed emergent patterns if cells were not allowed to also spontaneously differentiate at a low rate random rate . Although for the positive feedback rule this follows directly from the construction of the probability equations , the formation of strongly connected patterns was also not observed in the competing feedback rule without the inclusion of this low stochastic rate ( data not shown ) , suggesting this random rate is important to capture the experimentally observed transition patterns . The static structure model predicted small differences in the Oct4 expression kinetics of EBs with different cell seeding numbers should exist; however , the slight changes observed did not capture the full extent of the variation present in the experimental results . The inclusion of cell division ( with a faster division rate for pluripotent cells ) was also not able to explain the difference in the kinetics of this process . These results indicate that additional factors in the changing culture environment may modulate the kinetics of Oct4 loss in a size-dependent manner; hydrodynamic effects [60] , diffusion limitations and/or local chemical gradients may need to be taken into account for changing aggregates sizes [32] , [61] in order to reproduce the experimentally observed differences in differentiation kinetics . It is also possible that the rules chosen are not descriptive of our system with a fine enough resolution , thus explaining the ability to explain pattern formation characteristics but not kinetics . Furthermore , we constructed the model such that all cells have the same strength when affecting other cells , which assumes that cells convey the same amount of information regardless of the amount of cell connections or the amount of shared cell area , an assumption that may need to be refined as more detailed information about the nature of intercellular communication is included . Future developments will account for cell migration , local versus distal cell-cell communication and diffusion within the EB to investigate how these traits affect the physical microenvironment . The top-down modeling approach described in this study provides new insight into the spatial pattern development associated with differentiation of ESCs in 3D EB structures . Surprisingly , without explicitly modeling diffusion gradients or specific signal transduction mechanisms , features of temporal and spatial regulation were elucidated . Transitions in cell phenotype can be modeled as a flow of information between cells that is largely based on stochastic fate changes that are influenced by the cells' surrounding environment . The model was able to identify a simple rule paradigm that is biologically relevant and consistent with the current knowledge of stem cell regulatory processes [17] , [53] , [58] , [59] . A comparison of static and division models indicated that the proposed rule schemes are not significantly affected by cell division . Consequently , the proposed modeling technique developed thus far has demonstrated validity for exploring the propensity of various spatial patterns observed in EBs during the differentiation process . This model can readily be extended to investigate what factors influence differentiation into different germ layers , and eventually used to predict optimal niche and microenvironment organization for efficient stem cell maintenance and differentiation .
A murine embryonic stem cell line ( D3 ) transfected with an Oct4-GFP construct were used ( phOCT3-EGFP1; provided by Wei Cui , Ph . D . , Imperial College , London , UK ) . For this particular experiment these cells were used after several passages and splits , and thus did not show robust GFP activity under confocal or flow cytometry; thus immunostaining was necessary to visualize Oct4 expression . These cells were cultured in monolayer on 100 mm tissue culture plates coated with 0 . 67% gelatin in Dulbecco's modified Eagle's medium ( DMEM ) supplemented with 15% fetal bovine serum ( FBS ) ( Hyclone , Logan , UT ) , 2 mM L-glutamine ( Mediatech ) , 100 U/ml penicillin , 100 ug/ml streptomyocin , and 0 . 25 ug/ml amphotericin ( Mediatech ) , 1× MEM nonessential amino acis solcuiotn ( Mediatech ) , 0 . 1 mM 2-mercaptoethanol ( FisherChecmical , Fairlawn , NJ ) , and 103 U/ml leukemia inhibitory factor ( LIF ) ( Chemicon Internation , Temecula , CA ) . Cells were passaged every 2–3 days prior to reaching 70% confluence . Undifferentiated embryonic stem cells were dissociated from monolayer culture using 0 . 05% trypsin-EDTA solution ( Mediatech ) to obtain a single cell suspension and added to AggreWells ( Stem Cell Technologies ) [62] six well plate inserts to form six thousand EBs of either 250 ( 1 . 5 million cells/ml ) and 1000 ( 6 million cells/ml ) cell per EB . EBs were allowed to form in the wells for 20 hours , at which point they were removed and transferred into a rotary culture at 60 RPM [63] . EBs were re-fed every 2 days , and 75% of the spent medium was replaced with fresh medium at each exchange . EBs were cultured in this manner for the entire 7 day culture period . EBs were harvested at various time points and fixed for 45 minutes in 10% formalin . EBs were imaged using bright field microscopy via a 4× objective on an EVOS microscope . Three representative images were taken for each sample . Images were analyzed by using threshold , watershed , and image particle detection operations in ImageJ . EB radius was derived by computing the cross sectional area , approximating the EB as a circle , and calculating the radius accordingly . The circularity of the EBs was calculated by fitting an ellipse to the area , and taking the ratio of the minor and major axes . EBs were collected for staining and fixed in 10% formalin for 45 minutes . EBs were permeabilized for 30 minutes in 1 . 0% TritonX-100 , re-fixed in formalin for 15 minutes , and blocked in blocking buffer ( 2% bovine serum albumin , 0 . 1% Tween-20 in PBS ) for 3 hours . Samples were stained with a goat Oct4-antibody ( Santa Cruz ) overnight at 4°C . After three washes in blocking buffer , EBs were subsequently stained with a secondary donkey anti-goat Alexa Fluor 488 conjugated antibody ( 1∶200 Santa Cruz ) for 4 hours . Staining with Alexa Flour 546 Phalloidin ( 1∶20 Molecular Probes ) and Hoescht ( 1∶100 ) was performed concurrently for 25 minutes . Samples were washed , resuspended in blocking buffer , and imaged using a Zeiss LSM 510 Confocal Microscope using Ar , He , Ne and Chameleon lasers . A single image was taken at the top of the EB and at a depth of 25 µm into the EB . For each sample , 25 images were obtained . Spatial patterns of Oct4 expression were classified into six different categories , random , inside-out , outside-in , connected , differentiated , or undifferentiated . For an image to be classified as undifferentiated , 90% or more of the cells in the image had to positively express Oct4 . Conversely , for an image to be classified as differentiated only 10% of the cells could stain positive for Oct4 . If the number of positive Oct4 cells fell in between these two levels , the pattern of differentiation was classified as either random , inside-out , outside-in , or connected ( Fig . 4 ) . Inside-out patterns were characterized by differentiation in the middle of the EB , and undifferentiated cells on the outside . Conversely , outside-in patterns had differentiation on the outside and undifferentiated cells in the middle . Connected patterns were defined as multiple distinct connected regions of cells of the same state . Random patterns were classified as no identifiable pattern . For each time point 25 confocal images were analyzed . Pattern matching was performed on the output from the model as well as the experimental confocal images . Two blinded observers were used to classify the experimental confocal images . In the case of the in silico results , metrics such as the total number of differentiated cells , and average number of distinct cell clusters were used to aid in classifying patterns , with a total of 73 1000-cell aggregates and 66 250-cell aggregates analyzed . A cluster was defined as two or more cells of the same type and clusters were identified throughout the entire 3D aggregate . Rules-based modeling was carried out using probabilities to govern states changes . Linear , hyperbolic and hill-type probabilities have been previously examined in the context of stem cell differentiation and robustness and thus we used similar probability laws in this work [64] . All of these rules were designed to be functions only of the number of nearest neighbors to reduce complexity . For the “random” rule , a basal probability associated with the state change was set to 1% ( Fig . 4A ) . In the case of the “positive feedback” rule , the differentiation probability was influenced according to equation 1: ( 1 ) where β represents the number of nodes connected in the differentiated state , normalized by the total possible number of neighboring nodes ε , which for a face-centered cubic or hexagonal close-packed spherical packing arrangement is 12 ( Fig . 4B ) . In the case of the “competing feedback rule” rule the probability was determined according to equation 2: ( 2 ) where β represents the number of neighboring nodes in the differentiated state , and γ represents the number of neighboring nodes in the undifferentiated state ( Fig . 4C ) . This function produces a similar sigmoidal shape as the Hill function , but does not require the inclusion of the additional hill coefficient . Rules-based modeling was achieved using a Python language with the following freely available software packages: pyode , numpy , matplotlib , python imaging library ( PIL ) and vpython . Physical aggregation simulations of structure were performed using PyODE as the underlying physics engine . Results were plotted via the aid of numpy and matplotlib . The 2D aggregate slices were visualized using PIL . The 3D aggregate was visualized using vpython . Simulations were run until all of the cells had changed state or 500 time steps had elapsed . Cells were allowed to make fate decisions ever time step according to the probabilities outlined in the Rules-Based modeling section . The time-step cutoff was arrived at by taking the average number of time steps the “random” rule simulations ( as the “random” rule took the longest ) took to finish . Unless otherwise noted , 10 simulations were run for each rule condition . Simulations were run on an Intel Core i7 X980 3 . 33 GHz CPU with 12 . 0 GB of RAM . After aggregate sizes were determined , the number of cells in a spheroid was approximated by first determining the volume of the spheroid based upon known EB radii . Next we calculated the volume of an average cell using our average cell diameter of 6 . 6+/− . 3287 µm . By assuming a maximal close packed configuration for spheroids ( . 7408 ) , the volume of the aggregate was adjusted to contain the cells . Cell numbers were then arrived at by dividing the adjusted aggregate volume by the volume of a single spheroid . To calculate the growth rate equation 3 was applied between discrete time points . ( 3 ) This method produced growth rates over the first time course which closely matched the proliferation of mESC d3s in 2D . These growth rates could then be fit to equation 4 to determine doubling times which were used in the model for the different cell types . ( 4 ) Modeling of dynamic cell movement and vision was accomplished using custom C# code with the aid of XNA package for vector math and 3D visualization . Cells were modeled as rigid spheres connected by spring to depict cell-cell physical connections . A complete collision detection algorithm was used to resolve all possible collisions at each time step of the simulation . Simulations were run for a period of 144 hours ( 6 days ) , until 99 percent of the cells had changed state or until forty thousand total cells existed in the model . Cells were allowed to change fate instantaneously . The kinetics of the simulations were fit to model growth curves , thus the probabilities were given different weights to assure pattern formation was observed . In the case of the random and positive feedback rules , no weights were applied to the rules . However , in the case of the competing feedback rule a 0 . 01 weight was applied . 10 simulations were run for each condition . Simulations were run in parallel using on an Intel Core i7 X980 3 . 33 GHz CPU with 12 . 0 GB of RAM . The static and dynamic models were created with several key assumptions . With regards to the probability rules themselves , it was assumed that the rules were functions of the information from immediate neighbors . The rule were derived from previous literature which examined similar probability functions [64] . Also it was assumed that the cells could be expressed in a binary state based on previous work modeling differentiation as an all or none response [52] . For the static model the following assumptions apply: first , the probability functions are applied every time step as the kinetics of pattern formation were not the focus of these models . Second , neighboring cells were assumed to have the same amount of influence over each other regardless of the number of neighbors . That means that in the case where a cell only has one connection , this connection still conveys the same weight as any given connection in the rest to the network for conveying information . The final assumption for this model was that the cells change fate instantaneously as kinetics was again not a focus . For the dynamic model the following assumptions were used: first , it was assumed that the cells were not synchronized in terms of division . Second , cell division was symmetrical , meaning that cells only produced cells of the same type upon division . Third , cells were again assumed to have the same amount of influence over their neighbors regardless of the number of connections . Finally , the loss of pluripotency process was again modeled as an instantaneous transition . | Pluripotent embryonic stem cells can differentiate into all cell types making up the adult body; however , this process occurs in a complex three dimensional environment with many different parameters present that are capable of influencing cell fate decisions . A model that can accurately predict the strengths of factors influencing cell fate would allow examination of spatial and temporal patterns of cell phenotype . For this study , we focused on the earliest fate transition that occurs in 3D clusters of embryonic stem cells by monitoring the presence of a transcription factor ( Oct4 ) associated with stem cell pluripotency . After experimentally classifying patterns that arise en route to a fully differentiated aggregate in vitro , we constructed a computational model to deduce how stem cells integrate cues from their surrounding environment to give rise spatial patterns . We used a rules-based modeling approach in which information exchanged by cells with their nearest neighbors regulated cell fate decisions . This parsimonious model captured the spatial dynamics of early cell lineage commitment in a 3D multicellular structure . In combination with strategies to modulate cellular environments , our model provides a flexible tool for elucidating the extra- and intercellular interactions regulating spatially organized differentiation of stem cells in 3D . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"developmental",
"biology",
"computer",
"science",
"molecular",
"cell",
"biology",
"computer",
"modeling",
"stem",
"cells",
"cell",
"biology",
"embryonic",
"stem",
"cells",
"cellular",
"types",
"biology",
"computational",
"biology",
"stem",
"cell",
"lines",
"cell",
"differentiation"
] | 2013 | Spatial Pattern Dynamics of 3D Stem Cell Loss of Pluripotency via Rules-Based Computational Modeling |
Identifying drug-drug interactions ( DDIs ) is a major challenge in drug development . Previous attempts have established formal approaches for pharmacokinetic ( PK ) DDIs , but there is not a feasible solution for pharmacodynamic ( PD ) DDIs because the endpoint is often a serious adverse event rather than a measurable change in drug concentration . Here , we developed a metric “S-score” that measures the strength of network connection between drug targets to predict PD DDIs . Utilizing known PD DDIs as golden standard positives ( GSPs ) , we observed a significant correlation between S-score and the likelihood a PD DDI occurs . Our prediction was robust and surpassed existing methods as validated by two independent GSPs . Analysis of clinical side effect data suggested that the drugs having predicted DDIs have similar side effects . We further incorporated this clinical side effects evidence with S-score to increase the prediction specificity and sensitivity through a Bayesian probabilistic model . We have predicted 9 , 626 potential PD DDIs at the accuracy of 82% and the recall of 62% . Importantly , our algorithm provided opportunities for better understanding the potential molecular mechanisms or physiological effects underlying DDIs , as illustrated by the case studies .
Drug-drug interaction ( DDI ) is a significant cause of adverse drug reactions ( ADRs ) , especially in patient populations on multiple medications . A recent study indicated that medications were commonly used together in older adults , with nearly 1 in 25 individuals potentially at risk of a major DDI [1] . Approximately 70% of interactions are clinically relevant and contribute to the majority of ADRs [2] . DDIs occur when the pharmacologic effect of a given drug is altered by the action of another drug [3] , leading to unpredictable clinical effects . DDIs can be categorized into three types: pharmaceutical , pharmacokinetic ( PK ) , and pharmacodynamic ( PD ) [4] , [5] . Pharmaceutical interactions occur because of a physical or chemical incompatibility . A PK interaction occurs when one medication alters the absorption , distribution , metabolism , or excretion of another , changing the drug concentrations arriving at the target sites . PD interactions occur if one drug has an antagonistic , additive , synergistic or indirect pharmacologic effect on another . Current studies mainly focused on PK ( especially Cytochrome P450 enzymes ) DDIs and established experimental and simulation approaches to test for metabolic or transporter-based drug interactions [6] . However , a large number of DDIs cannot be explained at the PK or pharmaceutical levels and are supposed to be potential PD DDIs ( Figure S1 , Materials and Methods ) . Many of these interactions are not easily discernible because the endpoint is often a potentially serious adverse event rather than a measurable change in the concentration of the drug [5] . Typically , the potential PD DDIs were mainly based on sporadic cases reported during clinical trials . A number of severe PD DDIs are not identifiable in the early stage and result in great losses to human health . Thus far , the computational solutions to predict DDIs have used two distinct approaches . The first approach , termed similarity-based , predicted DDIs by measuring the similarity of drug information . As an example , Gottlieb et al [7] utilizes multiple drug-drug similarity measures to predict DDI . In this respect , many previous methods which were originally designed for inferring novel potential targets of drugs based on various types of data , such as structures [8] , targets [9] , indications [10] , side-effects [11] and gene expression profiles [12] , can also be used to infer drug interactions . The second approach is the knowledge-based approach that predicts DDI from scientific literature [13] , an electronic medical record database [14] and the FDA Adverse Event Reporting System [15] . However , both approaches suffer from several limitations , such as the necessity to distinguish drug classes and the inability to handle novel drugs for which limited reports exist [7] . More importantly , they seldom consider drug actions and their clinical effects in the context of complex biological networks . To ameliorate this situation , we adopted a network pharmacology strategy [16] , which considers drug actions and their clinical effects in the context of molecular network systems , and proposed an algorithm to systematically predict PD DDIs . Using known PD DDIs as golden standard positives ( GSPs ) , we demonstrated the superiority of our algorithm over previously published methods . The predictions also agreed with similar clinical side effects between the drugs , which was further incorporated with S-score to increase the prediction performance through a Bayesian probabilistic model . Importantly , our methods provided not only a comprehensive list of potential PD DDIs , but it also opportunities for further understanding of the molecular mechanism and physiological effect underlying DDIs .
To determine whether the network pharmacology strategy can be used to understand DDIs , specifically PD DDIs , we first investigated whether PD DDIs are reflected at the network level . We examined the distribution of the targets of drug pairs with known PD DDIs among the 1 , 249 FDA-approved drugs collected in DrugBank [17] in a protein-protein interaction ( PPI ) network from HPRD [18] . The connection for any possible drug pairs was measured by the minimum shortest path between their targets in the PPI network ( Figure 1A , Materials and Methods ) . Out of the 21 , 049 drug pairs that have minimum target distances of zero , 924 ( 4 . 4% ) were known PD DDIs ( Figure 1B ) , which represents a ∼6-fold enrichment compared with all possible drug pairs . This is expected because drug pairs with minimum distances of zero are those sharing at least one overlapping target , and these have been reported to have a high probability to form DDIs [9] . More importantly , we found that the smaller the minimum distance between two drugs' targets the more likely a PD DDI occurs ( Figure 1B ) , suggesting PD DDIs can be discerned at the PPI network level . In fact , the drug pairs with the minimum distance ≤3 already cover the majority ( >80% ) of the known PD DDIs ( Figure 1C ) . Overall , the average distance of known PD DDI targets is significantly shorter than the global average of possible drug pairs in the network ( P-value<2 . 2E-16 , Wilcoxon rank sum test ) . Based on the above observation , we designed a metric for systematically predicting PD DDIs by considering drug actions in the context of the PPI networks . First , drugs were mapped onto a PPI network based on their drug-target associations ( Figure 2A ) . Second , many drugs exert their therapeutic or adverse effects by interfering with tissue-specific molecular targets that are usually located in the same tissue where a pathological process occurs [19] . Therefore , we weighed the PPI in the network by Pearson's correlation coefficient ( PCC ) of their encoding genes' expression profile across 79 human tissues [20] ( Figure 2B ) . Then we defined a target-centered system for each drug , which includes drug targets and their first-step neighboring proteins in the PPI network ( Figure 2C ) . Finally , we defined a system connection score ( S-score ) to describe the connection between two target-centered systems in the PPI network as the following ( Figure 2D ) :where , and represent the mean , standard deviation and number of the cross-tissue expression PCC of edges connecting two drug-centered systems , respectively; represents the average PCC of all edges in the network as background . In addition , if two target-centered systems share a gene , an artificial edge with PCC of 1 is added between the two systems . Thus , S-score reflects the tightness of connection between two target-centered systems in the network , which not only depends on the number of edges connecting the genes in these two target-centered systems but also on the similarity in expression patterns across tissues . To evaluate our scoring scheme , we calculated S-scores for all possible drug pairs among the list of FDA-approved drugs . Using known PD DDIs collected in DrugBank as GSPs , we first evaluated the correlation between S-score and the likelihood that a PD DDI occurs . Indeed , the occurrence of PD DDIs decreased with decreasing S-scores among all possible drug pairs ( Figure 3A ) . Additionally , there was a highly significant correlation between S-score and the hits enrichment of GSPs ( R2 = 0 . 66 , P-value = 4 . 3E-52 ) ( Figure 3B ) . It indicated that the likelihood of a PD DDI to occur is high if the two drugs' targets are highly connected in PPI network and co-expressed in the same tissues . Next , we used receiver operating characteristic ( ROC ) curves to examine the performance of our algorithm . We compared our prediction with previously published methods ( Materials and Methods , Text S1 ) : ( 1 ) target overlap , connecting two drugs if share at least one target [9]; ( 2 ) target distance , connecting two drugs by their minimum distance of shortest path between targets on PPI network ( Materials and Methods ) ; ( 3 ) P-score , connecting two drugs by their side-effect similarities [11]; ( 4 ) C-score , connecting two drugs by their gene-expression signatures connectivity [12]; ( 5 ) indication overlap , connecting two drugs if they share a similar indication [10]; ( 6 ) text mining , connecting two drugs based on a co-occurrence scheme [13]; ( 7 ) TWOSIDES , a database of polypharmacy side effects for pairs of drugs mined from FDA Adverse Event Reporting System [15]; ( 8 ) INDI , a method predicted DDIs utilizing multiple drug-drug similarity measures [7] . Targets-based methods ( target overlap , target distance and S-score ) are better than those using indication , gene-expression signatures or side-effect similarities to connect drugs ( Figure 3C ) . Importantly , S-score , by integrating the information from drug-target associations , PPI network topology and cross-tissue gene co-expression , has the best performance ( Figure 3C ) . Interestingly , using different types of known DDIs as GSPs ( Materials and Methods ) , we found that S-score mainly predicted PD , but not PK or pharmaceutical DDIs ( Figure 3D ) . Using the DDIs recorded in DrugBank as GSPs , we also observed that our method outperformed previous methods ( Figure S2A ) . By using the drug-drug associations with medium text mining confidence score from the STITCH database [13] as another evaluation criterion , we also confirmed the robustness of S-score in predicting potential DDIs ( Figure S2B ) , even for our novel predictions which excluded the known DDIs in DrugBank ( Figure S2C ) . These results excluded the possibility that the performance of S-score was associated with biases of our semi-automatic text-mining method of classifying known DDIs into three types , and demonstrated the good performance of S-score is independent of the GSPs used . Expectedly , taking a negative set with a different size had a negligible effect on the result ( Figure S2D ) . To further validate our predictions , we examined the phenotypic effects of our predictions using published drug clinical side effect data [21] . Based on the observation that similar disorder phenotypes indicate overlapping molecular mechanisms [22] , we asked whether two drugs have similar clinical outcomes if they are highly connected in their target-centered systems ( Figure 2E and 2F ) . We measured the phenotypic connections between two drugs by their side-effect similarities ( P-score ) following a published algorithm , which was originally designed to infer novel potential targets of marketed drugs [11] . The drug pairs with high S-scores indeed had more similar phenotypes ( Figure 3E , P-value = 2 . 0E-72 , Wilcoxon rank sum test ) . Thus , S-score calculated using PPI network might partially explain the drug phenotypic overlap . To further increase the prediction performance , we integrated the evidences from S-score and P-score as a likelihood ratio ( LR ) using a Bayesian probabilistic model ( Figure 2G , Materials and Methods ) . As a result , we observed a clear improvement of prediction specificity and sensitivity ( Figure 3C ) . The area under the ROC curve ( AUC ) increased from 0 . 674 to 0 . 731 . In particular , for drug pairs with both evidences , the AUC of LR ( defined as LR ( S-score and P-score ) ) , approached 0 . 812 ( Figure 3C ) . We applied the algorithm to the FDA-approved drugs and generated a list of prioritized drug pairs where PD DDIs might likely occur . Overall , the list of 9 , 626 drug pairs with LR ( S-score and P-score ) >2 were 7 . 5-fold enriched for known PD DDIs against all possible drug pairs ( Table S1 ) , which represents an accuracy of 82% and a recall of 62% ( Materials and Methods ) . To further assess our novel predictions , we evaluated the potential side effects of our novel predictions against the TWOSIDES database [15] , which collected polypharmacy side effects for pairs of drugs from the FDA Adverse Event Reporting System ( Figure 2J ) . We observed a significant overlap between our novel predictions and TWOSIDES ( P-value<2 . 2E-16 , Fisher's exact test ) , where 27% of the novel predictions overlapped the list of TWOSIDES . The percentage approached 60% for our top 100 novel predicted drug pairs ( Table S1 ) . The prioritized list together with the available drug indication information , such as whether two drugs were likely co-used , can provide the rationale for which PD DDIs we should be mindful of during clinical trials or treatment . The most common drugs at the top of the prioritized list of potential PD DDIs were associated with tricyclic antidepressants ( TCA ) ( Table S1 ) , which are primarily used in the clinical treatment of mood disorders such as major depressive disorder ( MDD ) and dysthymia . It has been reported that patients taking antidepressants have more opportunities to experience DDIs , because antidepressants are often prescribed for months or years . In addition , patients with depressive disorders typically have comorbid symptoms that require administration of concomitant medications [23] . Although many of these drug interaction mechanisms remain unclear , it is recommended that concomitant therapy of TCAs should be used with caution considering the major clinical significance [24] , [25] . As an example , within the top 10 predicted DDIs , a potential interaction was predicted between two TCAs ( desipramine and trimipramine ) ( Table S1 ) . Such an interaction has been reported to increase the risk of additive QTc-prolongation and serious ventricular arrhythmias in DrugBank [17] . In our network model , the target-centered systems of these two drugs highly overlapped and connected with correlated cross-tissue gene expression ( Figure 4A ) , which is indicated by an S-score of 9 . 6 ( Student's t-test P-value<2 . 2E-16 , compared to all possible drug pairs ) . Interestingly , both of two drugs' target-centered systems are enriched in genes associated with the Gene Ontology “regulation of heart contraction” ( GO:0008016 ) ( P-value = 6 . 9E-5 and P-value = 1 . 6E-3 , respectively ) , which might help in explaining the molecular basis of the potential outcome of the concomitant administration of the two drugs . Our novel predictions together with the information from TWOSIDES provided opportunities for better understanding the potential molecular mechanisms or physiological effects underlying DDIs ( Figure 4B , Figure S3 and Text S1 ) . As an example , an interaction was predicted to exist between zonisamide and memantine ( Figure 4B ) . Zonisamide is a sulfonamide anticonvulsant approved for using as an adjunctive treatment of partial seizures in adults with epilepsy by blocking sodium and calcium channels , which leads to the suppression of neuronal hypersynchronization ( i . e . convulsions ) [17] . Memantine , an amantadine derivative used in the treatment of Alzheimer's disease , exerts its action through uncompetitive NMDA receptor antagonism , which protects against elevated concentrations of synaptically released glutamate in the brain of demented patients [17] . The two drugs do not have common targets , but do have similar cross-tissue expressions between their drug-centered systems ( S-score = 6 . 5 , P-value<2 . 2E-16 ) and similar side effects ( P-score = 73 . 5 , P-value<2 . 2E-16 ) . Although it has not been reported in DrugBank [17] , TWOSIDES recently reported that this drug pair has an significant association with the adverse event thrombocytopenia ( P-value = 1 . 36E-177 ) in the FDA Adverse Event Reporting System , which cannot be clearly attributed to the individual drugs alone [15] . Our analysis reveals that the genes in two drug target-centered systems are highly enriched in genes significantly highly expressed in the “Platelet” ( UP_TISSUE ) ( P-value = 8 . 8E-3 ) . Interestingly , such an interaction cannot be predicted based only on the knowledge of their drug targets as neither of the individual drug's target gene set is related to the thrombocytopenia symptom . Yet , consistent with their intended effects , emantine's targets are enriched for “N-methyl-D-aspartate selective glutamate receptor complex” ( GO:0017146 ) ( P-value = 1 . 4E-2 ) , which is involved in Alzheimer's disease [26] , while zonisamide's targets are enriched for “voltage-gated sodium channel complex” ( GO:0001518 ) ( P-value = 3 . 1E-4 ) , which is involved in pathological alterations in epilepsy [27] . Additional examples of novel predictions of the PD DDIs can be found in Figure S3 and Text S1 .
Despite the many methods previously applied to identify potential drug interactions from different aspects , these approaches have various limitations . To our knowledge , we for the first time , present an algorithm for systematically predicting PD DDIs by considering drug actions and their clinical effects in the context of complex PPI networks . The integration of various sources of information such as drug targets , network topology , cross-tissue gene expression correlations and side effect similarity indeed give rise to a better performance in predicting DDIs than those obtained with individual data sources . Finally , our network model provides opportunities for better understanding the potential molecular mechanisms or physiological effects underlying DDIs . However , like other computational-based techniques in this field , there still exists a gap between our scientific predictions in theory and clinical application . First , limited by the current knowledge of the molecular network as well as the robustness of the biological system itself , our prediction only provides the relative likelihood of the occurrence of a PD DDI . Second , as currently only a few types of data were used for prediction , the prediction power is bound to improve when integrated with more clinical data , if available , and complemented with recently published DDI prediction methods from different aspects . Last , the predicted potential PD DDIs are not necessarily always harmful but sometimes can also be beneficial [28] . Even though the current GSPs include only a small number of beneficial interactions , such interactions may occur through the same mechanism - overlapping network , in which case can be predicted by our method . With these further improvements , our method can be potentially applied in drug discovery and development , serving as an in silico systematic screen to provide a list of prioritized potential PD DDIs in a cost-effective manner or be applied to relabeling drug interaction warnings for marketed drugs . Our method can also reveal potential mechanisms or effects underlying DDIs and provide the necessary scientific evidence for further investigation of the drugs during clinical trials . These mechanisms could be valuable for rational poly-medication among existing drugs for new purposes to enhance beneficial drug combinations while avoiding harmful DDIs .
Drug information was downloaded from DrugBank database ( http://www . drugbank . ca/ ) on May 9 , 2011 . In DrugBank , a drug target is defined as “a protein , to which a given drug binds , resulting in an alteration of the normal function of the bound molecule and a desirable therapeutic effect” . In our further analysis , we mainly focused on the list of 1 , 249 FDA-approved drugs which include 4 , 776 associations with 1 , 289 targets . A PPI network , including 34 , 998 edges , was taken from Human Protein Reference Database ( HPRD; http://www . hprd . org/ ) [18] on Dec 7 , 2010 . To weight the edges in the network , we used PCC based on the pair-wise gene expression profiles in 79 human tissues [20] . To compare the prediction performance of our algorithm with previously published methods , we selected several representative methods in the field of DDI prediction [7] , [13] , [15] and also covered some approaches that were originally designed for inferring novel potential targets of drugs but can also be used to infer drug interactions [8] , [9] , [10] , [11] . To integrate the evidences from network system connectivity score ( S-score ) and drug phenotypic similarity score ( P-score ) , we used a Bayesian probabilistic model described in Xia et al [29] , where the Bayesian model has been proven to be particularly competent in predicting PPIs by integrating various evidences . The method has also been used to combine the different types of clues for predicting PPIs in a paper recently published [30] . Briefly , in the Bayesian probabilistic model , each score is automatically weighted according to its confidence level . The scoring schemes ( S-score , P-score ) were integrated as a likelihood ratio ( LR ) for drug pairs to be true positive DDIs versus true negative DDIs by multiplying from all the independent evidences as following:where LR ( i-score ) represents the likelihood ratio of evidence i-score . It relates prior and posterior odds according to the Bayes rule:where the terms ‘posterior’ and ‘prior’ refer to the condition before and after considering the evidence information i-score; the prior odds ( Oprior ) of finding the positive and negative hits can be can be calculated by considering the total number of GSP/GSN DDIs within all the possible drug pairs; the posterior odds ( Oposterior ) can be calculated by binning all possible drug pairs into discrete intervals according to the evidence i-score . We defined LR ( S-score and P-score ) for drug pairs with both evidences , while LR ( S-score or P-score ) for those with at least one evidence , respectively . Functional annotation analysis was performed using the DAVID web-server [31] . The datasets used in this paper and the core code in calculating the S-score were packaged and provided on our website http://www . picb . ac . cn/hanlab/DDI . | Drug-drug interaction ( DDI ) is an important problem in clinical practice . In this study , we developed a novel algorithm for systematically predicting pharmacodynamic ( PD ) DDIs through protein-protein-interaction ( PPI ) networks . We calculated a score to predict potential PD DDIs by integrating the information from drug-target associations , PPI network topology and cross-tissue gene expression correlations . The scoring system was validated by known PD DDIs and agreed with similarities in drug clinical side effects , which we further integrated to increase the prediction performance . Our approach not only outperformed previously published methods in predicting DDIs , but also provided opportunities for better understanding the potential molecular mechanisms or physiological consequences underlying DDIs . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"medicine",
"protein",
"interactions",
"drugs",
"and",
"devices",
"pharmacodynamics",
"genetics",
"and",
"genomics",
"pharmacology",
"biology",
"proteomics",
"drug",
"discovery",
"biochemistry",
"drug",
"research",
"and",
"development",
"genetics",
"genomics",
"drug",
"interactions",
"gene",
"networks",
"computational",
"biology",
"pharmacogenomics"
] | 2013 | Systematic Prediction of Pharmacodynamic Drug-Drug Interactions through Protein-Protein-Interaction Network |
Strongyloides stercoralis is a globally distributed nematode that causes diverse clinical symptoms in humans . Spain , once considered an endemic country , has experienced a recent increase in imported cases . The introduction of serology helps diagnosis and is currently replacing microbiological techniques in some settings , but its sensitivity is variable and can be low in immunocompromised patients . Diagnosis can only be confirmed by identification of larvae . Often , this “gold standard” can only be achieved in severe cases , such as disseminated S . stercoralis infection , or S . stercoralis hyperinfection syndrome , where parasite load is high . In addition , these clinical presentations are not well-defined . Our aim is to describe severe cases of S . stercoralis , their epidemiological profile , and their clinical details . An observational retrospective study of disseminated S . stercoralis infection , or hyperinfection syndrome . Inclusion criteria: aged over 18 , with a diagnosis of disseminated S . stercoralis infection , or hyperinfection syndrome , confirmed by visualization of larvae . Patients were identified through revision of clinical records for the period 2000–2015 , in collaboration with eight reference centers throughout Spain . From the period 2000–2015 , eighteen cases were identified , 66 . 7% of which were male , with a median age of 40 ( range 21–70 ) . Most of them were foreigners ( 94 . 4% ) , mainly from Latin America ( 82 . 3% ) or Western Africa ( 17 . 6% ) . Only one autochthonous case was identified , from 2006 . Immunosuppressive conditions were present in fourteen ( 77% ) patients , mainly due steroids use and to retroviral coinfections ( four HIV , two HTLV ) . Transplant preceded the clinical presentation in four of them . Other comorbidities were coinfection with HBV , Trypanosoma cruzi , Mycobacterium leprae or Aspergillus spp . All presented with digestive disorders , with 55 . 6% also presenting malaise . 44 . 4% of cases had fever , 27 . 8% skin complaints , and 16 . 7% respiratory or neurological disorders . One patient presented anemia , and one other nephrotic syndrome . Diagnosis was confirmed by identification of larvae in fresh stool samples ( n = 16; 88 . 9% ) , concentration techniques ( n = 6; 33 . 3% ) , larval culture ( n = 5; 29 . 4% ) , or digestive biopsies ( n = 8; 44% ) . S . stercoralis forms were identified during necropsy in one case . In addition , ten ( 55% ) had a positive serology . All the cases were treated with ivermectin , six ( 33% ) also received albendazole and one case received thiabendazole followed by ivermectin . All needed inpatient management , involving a mean hospitalization stay of 25 days ( range 1–164 ) . Two cases received intensive care and eventually died . Only eighteen cases of disseminated S . stercoralis infection/hyperinfection syndrome were identified from the 15-year period , most of which were considered to have been imported cases . Among those , immunosuppression was frequent , and mortality due to S . stercoralis was lower than previously described .
Strongyloidiasis is an infection caused by Strongyloides stercoralis . Other Strongyloides spp . include S . fuelleborni , which infects some non-human primates and may cause self-limited disease in humans . S . stercoralis is widely distributed , affecting up to 370 million people worldwide according to recent estimates [1 , 2] , and some experts claim that its prevalence is increasing around the globe [1 , 3] . Spain had an estimated prevalence of 0 . 2% in some rural areas during the first half of the last century [4] , especially where wet crops , such as rice , and animal tracks were common . The epidemiology of the disease in Spain changed notably in the early 80s owing to economic growth , the abandonment of traditional farming techniques and the mechanization of agriculture , along with an improvement in sanitation networks in rural areas . Although , some reports suggest that such transmission may still be occurring , [5 , 6] , all these advances have led to autochthonous cases becoming rare [5] . Economic growth brings both population ageing and an increase in the availability of diagnosis and treatment for chronic conditions , leading to a higher rate of comorbidities and also immunosuppressions in which severe cases might appear [6] . Recently published data from our group have shown a tenfold increase in hospital admissions where S . stercoralis is involved over the last decade [7] . This parasite undergoes a fascinating life-cycle that alternates between a free-living cycle in moist soil and a parasitic cycle in the host . The latter may last for decades , since the worm can replicate in the host and cause repeated autoinfections [8] . Parasite transmission to humans occurs during its free-living cycle through direct skin contact with previously contaminated soil containing rhabditiform larvae , although transmission through organ transplant has also been described [9 , 10] . Therefore , poor sanitary conditions facilitate parasite transmission . Strongyloidiasis is an indolent infection in most cases , although it can lead to mild digestive , skin or pulmonary symptoms . In some cases , usually due to comorbidities such as immunosuppression , replication of larvae can increase and they can disseminate into other tissues causing acute severe conditions [9] . These life-threatening forms of the disease comprise hyperinfection syndrome and disseminated strongyloidiasis . Both clinical pictures display a variety of symptoms , often with significant digestive involvement , that can produce paralytic ileus . In addition , larvae migration to the lungs can lead to hemoptysis and respiratory distress , and invasion of the brain can present as meningoencephalitis . Translocation of bowel bacteria causing sepsis and meningitis is often seen along with these findings . In both presentations , eosinophilia is an unfrequent finding and mortality can be as high as 62 . 7% [9] . Unfortunately , confirmation of diagnosis relies on visualization of parasite forms , which is unlikely with low parasite loads . Nevertheless , hyperinfection and disseminated forms are , by definition , situations where parasite load is high , so diagnosis in these cases is often confirmed . As a Neglected Tropical Disease , strongyloidiasis is often forgotten about even where it is most common , and more so where it was once prevalent , but is no longer . Some authors state that S . stercoralis hyperinfection syndrome is in fact an emerging disease , and point to a lack of awareness among health-care professionals in non-endemic areas [11] . In particular , the clinical presentation of disseminated strongyloidiasis , or hyperinfection syndrome , is poorly defined in the literature , often based solely on case reports [6 , 9] . Our aim is to define these severe forms of the disease as seen in our context .
This work was performed in accordance with the ethical standards laid down in the Declaration of Helsinki as revised in 2013 . The study protocol was approved by the Ethical Review Board of Vall de Hebron University Hospital ( Barcelona , Spain ) with the assigned code PR_AG_03–2016 . Since this was a retrospective observational study , our Institutional Review Board accepted to proceed to data compilation and analysis with no previous informed consent obtained from the participants . All clinical and epidemiological data were anonymized .
Of the ten reference centers which participated in the study , two found no cases of disseminated strongyloidiasis nor hyperinfection syndrome , while the eight remaining centers were able to recover cases from their clinical records for the period defined . From the years 2000–2015 , eighteen cases were identified . The majority were male ( n = 12 , 66 . 7% ) , which implies a sex ratio of 2:1 , and the median age was 40 ( range 21–69 ) . Most of them were foreigners ( n = 17 , 94 . 4% ) , mainly from Latin America ( n = 14 , 82 . 3% ) or Western Africa ( n = 3 , 17 . 6% ) . Of these patients , fourteen had information about the length of their period of residence in Spain , the median period being 7 . 21 years ( range 2 months-31 years ) , and in two cases , it was possible to identify a recent travel history involving a visit to friends and relatives in their native country , respectively six and eleven months before the onset of symptoms . Further epidemiological details are provided in Table 1 . Only one autochthonous case was identified , that of a 40-year-old construction worker from the Canary Islands with no travel history outside of Spain . He reported having walked barefoot in mud in a known endemic area in mainland Spain , and later developed hyperinfection syndrome after induced immunosuppression for a renal transplant . Most of the cases fulfilled the criteria for hyperinfection syndrome ( n = 17 , 94 . 4% ) , and only one case of dissemination was identified and confirmed by necropsy ( case 3 ) . This allowed for identification of S . stercoralis in skin , lungs , bowels , brain , kidneys and lymph nodes . The clinical details of each patient are specified in Table 2 and summarized below . Immunosuppressive conditions were identified in fourteen ( 77% ) patients and were mainly due to prolonged steroids use ( n = 9 , 50% ) , followed by retroviral coinfections ( four HIV , two HTLV ) . All the HIV cases were severely immunosuppressed , with CD4 counts of 8 , 179 , 10 and 22 at diagnosis ( cases 5 , 8 , 9 and 16 , respectively ) , and with detectable viral loads ranging from 390 , 000 to 4 . 300 , 000 copies/mL . In case 16 , S . stercoralis severe infection was diagnosed along with HIV infection , while the others developed S . stercoralis hyperinfection syndrome within three months of starting antiretroviral therapy ( ART ) . Of note , cases 5 and 9 had also received high dose steroids for cerebral toxoplasmosis before the onset of severe strongyloidiasis symptoms . Transplant preceded the clinical presentation in three patients , case 3 having received an allogenic blood stem cell transfer , and cases 7 and 13 having received a renal transplant . Case 2 developed hyperinfection syndrome after a renal transplant . Every transplanted case was also receiving steroids . A further two patients were under treatment with steroids; one for Sjögren syndrome , and the other for a type-2 lepromatous reaction ( cases 11 and 17 respectively ) . Other comorbidities were coinfection with Trypanosoma cruzi in three cases , hepatitis B virus in two cases , and Mycobacterium leprae and Aspergillus spp . in one case each . All presented with digestive disorders . Ten ( 55% ) patients presented with malaise . Eight patients ( 44% ) had eosinophilia , as defined above , when strongyloidiasis was diagnosed; out of these , five were considered immunosuppressed . Of note , two out of the four HIV patients presented with eosinophilia , and case 16 ( who was ART naïve at diagnosis ) had an increase in his absolute and relative eosinophilia after starting ART , as happened with case 17 after steroid therapy was discontinued; Eosinophil count ranged from 191 eosinophils/μL ( 5 . 3% ) to 11 . 200 eosinophils/μL ( 65 . 5% ) . Eight ( 44 . 4% ) cases had fever , five ( 27 . 8% ) had skin complaints , four ( 22 . 2% ) presented with respiratory complaints and three ( 16 . 7% ) with neurological disorders , one patient presented anemia , and one presented nephrotic syndrome . Diagnosis was confirmed by identification of larvae in every case . Nine out of ten participating centers used Ritchie’s concentration technique and one obtained larvae after concentration with ethyl acetate . Larvae were cultured with charcoal culture , Harada-Mori filter paper or blood agar plate . Every case had a large number of filariform larvae in fresh stool samples ( n = 16; 88 . 9% ) concentration techniques ( n = 6; 33 . 3% ) or larval culture ( n = 5; 27% ) . In some patients , S . stercoralis adult worms , eggs and larvae were identified in digestive and extra-digestive biopsies ( n = 8; 44% ) , often described with an eosinophilic infiltrate around them . S . stercoralis forms were identified during necropsy in one case . In addition , ten ( 55% ) had a positive serology , not performed in the remaining eight patients as the parasite had already been identified otherwise . All the cases were treated with ivermectin adjusted to 200 mcg per kg a day , with the duration of treatment ranging from two to ten days . Three patients receiving ivermectin for two days had this regimen repeated one or two weeks later . Furthermore , six ( 33 . 3% ) also received albendazole 400 mg before or concomitantly with ivermectin . One patient was treated with thiabendazole 1750 mg for three days in addition to a single dose of ivermectin . All needed inpatient management with a mean hospitalization stay of 25 days ( range 1–164 ) . Two cases ( 11 . 1% ) needed intensive care and eventually died .
We found eighteen cases of severe S . stercoralis infection during the 15-year period studied in four different regions of Spain , most of which were imported , mainly from Latin America . The male to female ratio was 2:1 , and the majority of patients were young adults . Immunosuppression frequently preceded the onset of symptoms , and mortality was lower than hitherto described , with a rate of 11 . 1% . Robson et al . sought cases of S . stercoralis hyperinfection syndrome diagnosed in the United Kingdom after the Second World War to our date and described the largest series outside endemic areas [13] . Other recent studies on this topic are case reports , or hospital case-series describing any type of strongyloidiasis regardless of its severity [6 , 14–16] . So ours is the biggest case series compilation outside endemic areas after that from Robson et al . , who described severe cases over six decades . One of the strengths of this study is the circumstances of the patients; since they were attended in reference centers , we have been able to fully describe the clinical picture and outcome of severe cases when all health-care resources are available in one place . However , this is also a disadvantage , as our data are not translatable to other clinical settings , nor to endemic areas . A further limitation of our study stems from the use of clinical records , as at times information was lacking due to the fact that it relied on the health worker’s thoroughness in registering data . For this reason , detailed information on exposure to risk factors was not available . The age , sex and origin of the participants match the epidemiology described in previous hospital series in our country [6 , 14–16] , probably reflecting the epidemiology among our immigrant population . Clinical presentations are in accordance with those seen in previous case reports of severe strongyloidiasis , with a predominance of digestive tract disorders and general physical malaise . Nevertheless , these clinical symptoms might be undistinguishable from a complicated chronic S . stercoralis infection without microbiological evidence of a high number of larvae . Eosinophilia was a relatively frequent finding ( n = 8 , 44% ) for such a population , since it is less frequently observed when cellular immunity is compromised . Of note , two of our patients with HIV developed eosinophilia once they started ART , and another presented eosinophilic infiltrate in a duodenal biopsy , findings which are in accordance with those described in other series [15 , 17] . Regarding comorbidities , a recent systematic revision of severe strongyloidiasis cases described a large number as having received steroids before the onset of symptoms ( 67% ) , a 15% HIV coinfection rate , and a further 11 . 5% of cases being transplant recipients [9] . In our series , immunosuppression could be identified in two-thirds of the cases , predominantly among those receiving steroids and oncological chemotherapy , but four patients seemed to have no factor which would have triggered a massive increase in S . stercoralis larvae . Serology is a sensitive tool although not highly specific . However , sensitivity can decrease to 42 . 9% in immnunocompromissed patients [18] . This technique was available in every center of our group . Since all our cases were confirmed with microbiological evidence , serology was only requested in ten of them . Out of these , six presented an immunosuppressive condition , raising the highest sensitivity described in such cases ( 100% ) , but we believe these numbers are not significant . Serology has also been recently proposed as a tool to monitor treatment success [16] . Unfortunately , it was not systematically requested during our patients’ follow-up . Recent research into screening strategies for imported diseases found a significant rate of intestinal parasites among those awaiting treatment for oncohematological malignancies [19] . Steroid treatment and immunosuppressive drugs provided prior to and after transplants were both attributable causes of severe strongyloidiasis in half of the patients in our series , since S . stercoralis infection was thought to have been present before these therapies were started . Nevertheless , other recent reports have also described S . stercoralis infection in donors for both solid organ and stem cell transplants [20 , 21] . Since the origin of the donors is unknown for the four patients who had received a transplant , they cannot be ruled out as a possible source of infection . Some expert groups in tropical diseases have found rates of up to 18 . 4% of S . stercoralis infection in HIV patients coming from endemic areas , suggesting that this coinfection may be common [22] . Among those infected with HIV in our series , the development of severe strongyloidiasis could be attributed to HIV infection alone in only one patient ( case 16 ) , while a further three might have developed S . stercoralis severe infection as a manifestation of an immune reconstitution phenomenon , as described in the literature [23–25] . However , cases 5 and 9 were also receiving high-dose steroids as part of the treatment for a cerebral toxoplasmosis , which probably played a role in triggering a higher larval load . An evidence-based guideline has been published during the development of this work . It recommends combination therapy with ivermectin and albendazole as the treatment of choice for severe cases [26] , which was the empirical treatment given to 33% of our cases . The different treatment strategies used during our study period were chosen according to the published evidence at the time and drug availability , meaning ivermectin was not always present at the center when the patient arrived and albendazole or thiabendazole had to be launched initially , while awaiting ivermectin . Nevertheless , some drug regimens had to be repeated weeks apart , in some cases lasting up to months , since no parasitological and/or clinical improvement was observed with single dosages , probably reflecting prepatent autoinfections arising due to the increased larvae replication . Buonfrate et al . found that all the patients suffering from severe strongyloidiasis who received ivermectin survived [9] . This treatment is the current drug of choice for the treatment of S . stercoralis infection and was given to every patient in our series [27 , 28] , what might explain the relatively low mortality described here ( 11 . 1% ) . In addition , they were managed in reference centers where the disease was suspected and diagnosed early , and where the coverage of concomitant sepsis with broad-spectrum antibiotics or intensive care were assured . Raising awareness about the disease among populations-at-risk and healthcare professionals is strongly recommended [26] . During a S . stercoralis and Trypanosoma cruzi screening campaign , performed in 2016 among Latin-American immigrants in Alicante , Spain , it was found that 92 . 2% of participants ( 119/129 ) had never heard of strongyloidiasis , including none of the ten participants who had a positive S . stercoralis serology ( personal communication ) . Moreover , a questionnaire about five Neglected Tropical Diseases completed by students from Madrid in their final year of Medicine , revealed that less than 18% of the students ( 18/103 ) ‘passed the exam’ on strongyloidiasis , this being one of the most worrying results [29] . We conclude that screening for strongyloidiasis should be mandatory for HIV patients , as well as for both transplant recipients and donors coming from endemic areas . Infection should also be ruled out in those diagnosed with HTLV-1 infection , and ideally before the onset of steroid treatment . At the time this paper was under revision , a panel of experts published an evidence-based guideline supporting this recommendation with a Ia grade [30] . It is clear that disease outcomes improve when clinicians are aware of the infection and ivermectin supply is available for patients who require it . | Strongyloides stercoralis is a globally distributed worm . It has a free living cycle in wet moist soils , and an autoinfecting cycle affecting humans in their lungs , bowels and skin . Strongyloidiasis is the name of the infection caused by S . stercoralis and it can vary from an indolent state , with no symptoms at all , to a severe clinical condition if the adult worms reproduce and disseminate into the body tissues . This second clinical picture usually presents when the immune system of the host is altered for any reason , most likely other immunosuppressive infections or treatments . Severe clinical conditions are usually confirmed when an increased number of S . stercoralis larvae are found in body tissues , which leads to a broad spectrum of clinical symptoms that are usually fatal without treatment . These conditions are both uncommon and not well defined . We present the experience of eight referral centers in Spain regarding the management of severe strongyloidiasis . We found 18 cases diagnosed over 15 years , the second largest series outside endemic areas , and compiled their epidemiological , clinical and outcome data . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"hiv",
"infections",
"urinary",
"system",
"procedures",
"medicine",
"and",
"health",
"sciences",
"organ",
"transplantation",
"european",
"union",
"tropical",
"diseases",
"geographical",
"locations",
"parasitic",
"diseases",
"surgical",
"and",
"invasive",
"medical",
"procedures",
"developmental",
"biology",
"renal",
"transplantation",
"neglected",
"tropical",
"diseases",
"europe",
"transplantation",
"infectious",
"diseases",
"blood",
"and",
"lymphatic",
"system",
"procedures",
"spain",
"eosinophilia",
"life",
"cycles",
"hematology",
"people",
"and",
"places",
"helminth",
"infections",
"hiv",
"diagnosis",
"and",
"management",
"diagnostic",
"medicine",
"stem",
"cell",
"transplantation",
"cell",
"transplantation",
"strongyloidiasis",
"biology",
"and",
"life",
"sciences",
"viral",
"diseases",
"soil-transmitted",
"helminthiases",
"larvae"
] | 2018 | Management of severe strongyloidiasis attended at reference centers in Spain |
Understanding the principles governing axonal and dendritic branching is essential for unravelling the functionality of single neurons and the way in which they connect . Nevertheless , no formalism has yet been described which can capture the general features of neuronal branching . Here we propose such a formalism , which is derived from the expression of dendritic arborizations as locally optimized graphs . Inspired by Ramón y Cajal's laws of conservation of cytoplasm and conduction time in neural circuitry , we show that this graphical representation can be used to optimize these variables . This approach allows us to generate synthetic branching geometries which replicate morphological features of any tested neuron . The essential structure of a neuronal tree is thereby captured by the density profile of its spanning field and by a single parameter , a balancing factor weighing the costs for material and conduction time . This balancing factor determines a neuron's electrotonic compartmentalization . Additions to this rule , when required in the construction process , can be directly attributed to developmental processes or a neuron's computational role within its neural circuit . The simulations presented here are implemented in an open-source software package , the “TREES toolbox , ” which provides a general set of tools for analyzing , manipulating , and generating dendritic structure , including a tool to create synthetic members of any particular cell group and an approach for a model-based supervised automatic morphological reconstruction from fluorescent image stacks . These approaches provide new insights into the constraints governing dendritic architectures . They also provide a novel framework for modelling and analyzing neuronal branching structures and for constructing realistic synthetic neural networks .
Neuronal circuits are composed of a large variety of branched structures – axons and dendrites – forming a highly entangled web , reminiscent of a stochastic fractal [1] . Despite this apparent chaos , more than a century ago Ramón y Cajal was able to extract order from this neuroanatomical complexity , formulating fundamental anatomical principles of nerve cell organization [2] . Cajal described three biological laws of neuronal architecture ( Chapter V , p . 115–125 , in [2] ) : optimization principles for conservation of space , cytoplasm and conduction time in the neural circuitry . These principles helped him to classify his observations and allowed him to postulate a wide variety of theories of functionality and directionality of signal flow in various brain areas . In the meantime , many of these ideas have been substantiated by applying more rigorous statistical analysis: circuitry and connectivity considerations as well as simple wire-packing constraints have been shown to determine the statistics of dendritic morphology [3]–[5] . It has also been shown mathematically that the specific organization and architecture of many parts of the brain reflect the selection pressure to reduce wiring costs for the circuitry [6]–[9] . In parallel , the development of compartmental modelling techniques based on the theories of Wilfrid Rall have highlighted the importance of a neuron's precise branching morphology for its electrophysiological properties [10] , and have shown that dendrites can play an important role in the computations performed on the inputs to the cell [11] , [12] . In fact , requirements for highly selective connectivity [13] , [14] , coherent topographic mapping , sophisticated computation or signal compartmentalization at the level of the single cell [15] and the network could all contribute to this observed intricacy of brain wiring . These two lines of investigation raise the question as to whether computation plays the determining role in shaping the morphological appearance of neuronal branching structures . Alternatively , the simple laws of material and conduction time preservation of Ramón y Cajal could have more influence . Using computational techniques it has become possible to construct synthetic neuronal structures in silico governed by the simulation of physical and biological constraints [1] , [16]–[21] . In two recent papers [19] , [20] , we derived a growth algorithm for building dendritic arborisations following closely the constraints previously described by Ramón y Cajal . The algorithm builds tree structures which minimize the total amount of wiring and the path from the root to all points on the tree , corresponding to material and conduction time conservation respectively . Synthetic insect dendrite morphologies were faithfully reproduced in terms of their visual appearance and their branching parameters in this way . Here we explore the algorithm's general applicability and its potential to describe any type of dendritic branching . If the algorithm is sufficient to accurately mimic the essential structure of neuronal circuitry we can resolve the relative importance of computation and wiring constraints in shaping neuronal morphology . We can then claim that Ramón y Cajal's laws are sufficient for shaping neuronal morphology . Specific computation will then only play a subordinate role in determining a neuron's branching pattern . We show here that while Cajal's laws do represent a strict constraint on neuronal branching , a neuronal morphology has a certain freedom to operate within these constraints . Firstly , by adjusting the balance between the two wiring costs , a dendrite can efficiently set its electrotonic compartmentalization , a quantity attributable to computation . Secondly , the density profile of the spanning field in which a dendrite grows determines its shape dramatically . Thirdly , a few weaker constraints such as the suppression of multifurcations , the addition of spatial jitter or the sequential growth of sub-regions of a dendrite are helpful for reproducing the dendritic branching patterns of particular preparations . These additional constraints might shed light on further functional , computational , developmental or network determinants for certain dendritic structures , and more of these will follow when applying our method to many more preparations . Moreover , the simple principles presented in this study can be used to efficiently edit , visualize , and analyze neuronal trees . Finally , these approaches allow one to generate highly realistic synthetic branched structures , and to automatically reconstruct neuronal branching from microscopy image stacks .
Before generating complex neuronal morphologies , a simple formalism is required to compare and assess natural and synthetic neuronal trees . We derive such a formalism from graph theory: a neuronal tree is a graph which connects a set of labelled nodes via directed edges away from a root labelled “1” ( Figure 1A ) . This distinct directionality is useful since properties describing a neuron's branching typically relate to the root of the tree , e . g . the branch order which increases after each branch point away from the root . In general , the graph describing a specific neuronal tree should be entirely unique in order to be used to compare two trees topologically ( their branching properties ) or electrotonically ( their functional properties ) . To achieve this , a unique labelling of the nodes is required . We constrain labelling by imposing a hierarchical order ( node label values increase with distance from the root ) , continuous labelling within sub-trees ( see for example nodes “6” , “7” , “8” and “9” , which form a sub-tree , Figure 1A ) and a topological sorting in which at any branch point the sub-tree with a higher topological depth is labelled first ( see Figure 1A and “label sorting” section in Methods ) . Apart from requiring unique labelling , a unique representation of a tree requires its nodes to be precisely distributed along its geometry . The process of manual reconstruction assigns node locations in an arbitrary manner ( Figure 1B , original tree ) . We introduce a process we term resampling , in which nodes are redistributed on the same tree structure , assigning homogeneous inter-nodal distances ( Figure 1B shows 10 and 20 µm resampling; see “resampling” section in Methods ) . When node labelling and distribution are attributed in a distinct manner , a unique representation of the tree is provided . By rearranging the node locations of a sample tree ( Figure 1C , sample tree ) based on its label order while preserving segment lengths , one obtains a unique electrotonic equivalent ( Figure 1C , equivalent tree; see “Electrotonic equivalent” section in Methods ) . Dendritic structure can then also be described by a single unique adjacency matrix , which indicates for each node its direct parent node ( Figure 1D , adjacency matrix ) . Consistent with the adjacency matrix , a matrix containing the current transfer from any node to any other in the tree is a good representation of its electrotonic properties ( Figure 1D , electrotonic signature ) . As a result of the continuous node labelling within sub-trees , electrotonic compartmentalization expresses itself as square sub-regions with high reciprocal current transfer . Because of the unique graphical representation , the electrotonic signature is independent of the reconstruction procedure . Note that simplified tree structures , which preserve the electrotonic compartmentalization , can be obtained by a coarser resampling ( Figure 1D , electrotonic signature of the same tree as in Figure 1B with 20 µm sampling ) . Computing current flow in a corresponding compartmental model is much faster since the number of nodes is decreased drastically ( from 297 to 39 in the example of Figure 1D ) . Finally , a simple and unique one-dimensional string can be used to describe the topology of the tree entirely , which we term the “topological gene” . In this string each branch is described in the order of its node labels by its length value followed by a “B” if the branch ends in a branch point or by a “T” if it ends in a termination point . The “topological gene” can be displayed as a sequence of green ( “B” ) and black ( “T” ) pieces ( Figure 1E ) . If diameter values for each node are known , the electrotonic signature can be reconstructed solely from this one-dimensional string , since segment lengths and topology are conserved . In order to incorporate Cajal's hypotheses about wiring optimization in our theoretical description of a neuronal tree , we implemented optimization procedures known from graph theory . This approach was previously shown to be successful for generating synthetic dendritic structures of fly interneurons [19] , [20] as well as recently for neocortical axons [22] . We now generalize it to more arbitrary neuronal geometries . Figure 2A exemplifies the general approach of assembling a set of unconnected carrier points to such an optimized graph . A greedy algorithm based on the minimum spanning tree algorithm [23] starts at the root with an empty tree and connects unconnected carrier points ( red dots ) one by one to the nodes of the tree ( black dots ) . At each step , the unconnected carrier point which is the point closest to the tree according to a cost function connects to the node in the tree to which it is nearest . The distance cost is composed of two factors: 1 ) A wiring cost represented by the Euclidean distance between the carrier point and the node in the tree ( red dashed lines show three sample segment distances for point P ) ; this quantity loosely corresponds to the material conservation constraint by Cajal; 2 ) A path length cost of the path along the tree from the root ( large black node ) to the carrier point; this quantity is consistent with the conduction time conservation constraint by Cajal . In the example here , even though P is closer to node 5 in Euclidean terms , the additional cost of path length ( adding node 5 on the path ) might tip the balance in favour of node 4 . A balancing factor bf , which weighs these two costs against each other in the cost function ( total cost = wiring cost +bf · path length cost ) , represents the one and only parameter of the model . Figure 2B illustrates the approach for neuronal trees grown on homogeneously distributed random carrier points in a circular envelope when the root is located at its centre . Since the two constraints ( minimizing wiring and minimizing path length to the root ) are weighted according to the balancing factor bf determining the contribution of the second constraint , the synthetic trees range along the dimension of that parameter from a pure minimum spanning tree , which grows in a wide spiral , to a purely stellate architecture ( Figure 2B , from left to right ) . In the following , we will apply this method of creating optimized graphs to reproduce morphologies in various neuronal preparations . The main effort will be to obtain an adequate set of carrier points for the application of the algorithm; this will prove to depend strongly on the density profile of the spanning field in the respective geometries . When additional constraints will be required in generating neurons in specific brain areas , this will provide clues pointing to actual computational or functional features of neuronal morphology . Whereas our previous work was limited to insect dendrites [19] , [20] , here we explored whether the algorithm is also able to reproduce a variety of neuronal structures . We first investigated the simple case of a planar neuron: the starburst amacrine cell of the mammalian retina . Its root is invariably located at the centre of a circular planar structure ( data from [24]; Figure 3A ) . This arrangement provides a common geometrical context for these cells . In order to best generate synthetic starburst amacrine cell-like neurites , random carrier points were distributed according to a ring-shaped density function around the centre in the root , limited by a simple circular hull ( Figure 3B ) . The locus of increased density most likely corresponds to the area where an increased number of connections is being made in the real cell , with directional selectivity probably being computed there [25] , [26] . Figure 3C demonstrates that this process successfully generates a synthetic neurite . The right balance between the two optimization constraints plays a crucial role , as is evident from a synthetic tree grown with a different balancing factor ( bf = 0 . 2 , Figure 3D ) . An appropriate balancing factor was determined by quantitatively comparing total cable length , mean path length to the root and number of branch points to the original real tree ( Figure 3E ) . Using the corresponding balancing factor resulted in realistic distributions of branch order and path length values as well as a realistic Sholl plot [27] , which counts the number of intersections of the tree with root-centred concentric spheres of increasing diameter values ( Figure 3F–H ) . The starburst amacrine cell neurite required a higher bf than did the insect dendrites ( 0 . 6 versus 0 . 4 , see [19] , [20] ) . Additionally , suppressing multifurcations improved the growth process ( compare Figure 3CD with Figure 2B ) . This was generally beneficial for all neurons studied here , and might reflect a constraint for the underlying developmental growth process . To better reproduce the appearance of reconstructions of real neurons , spatial jitter was added in all cases in the form of low-pass filtered spatial noise applied directly on the coordinates of the nodes in the resulting tree . Note that homogeneous noise application was only possible after the tree was resampled to a fixed segment length . Spatial noise in real reconstructions is partly due to fixation ( e . g . shrinkage or reconstruction artefacts ) and should therefore not necessarily be reproduced by the synthetic morphologies . However , wriggly paths in neuronal branching , corresponding to a spatial jitter along the branches , can be a result of obstacle avoidance and therefore can be associated with space packing issues [3] , relating to the third law described by Ramón y Cajal . In this study , however , we do not model volumetric optimization or space packing of other neuronal and non-neuronal structures in the tissue . We thus simply note here that in order to fully reproduce starburst amacrine cell reconstructions , multifurcations were suppressed and spatial jitter was added . We next studied dendrites of hippocampal granule cells , which fill a three-dimensional volume rather than a plane ( template data from [28]; see sample cells in Figure 4A ) . We first centred the original cell reconstructions on the soma location and rotated them manually in all three dimensions to produce axial symmetry with respect to each axis . Then , the dendrites were scaled to the average limits over the population of real morphologies for each of the three dimensions . Surprisingly , the spanning fields overlapped delineating again a common context for all cells ( Figure 4B ) . A geometric approach to describing the envelope of the dendrites is to intersect an elliptical cone with a sphere whose centre lies outside beyond the tip of the cone along the cone's central axis . The density profile of topological points ( branch and termination points only ) seemed to increase close to the origin of the volume and again further out at the rim ( Figure 4B ) . Growing synthetic trees on random carrier points distributed according to this type of density profile within the constructed cone-like volume resulted in realistic dendritic structures ( see examples in Figure 4C ) . Again , altering the balancing factor resulted in significant changes in branching behaviour ( Figure 4D ) . This could be used to determine an appropriate balancing factor , which was higher for these cells . For comparison , full distributions of branching statistics are shown for synthetic granule cells and real counterparts in Figure 4E–G in analogy to Figure 3F–H . The two cases indicate that adequately balancing the costs for wiring with the costs for path length distances to the root is crucial to describing the dendritic morphology . The two cases described in Figures 3 and 4 used a geometrical construction to produce the density profile from which the carrier points were obtained , i . e . a ring-like density confined by a circle or a bimodal density profile within a volume obtained by intersecting a cone with a sphere . We next tried to generalize the approach taken with hippocampal granule cells to all neuron types . We derived context and spanning fields for a wide variety of existing cell types using a common feature that we observed in hippocampal granule cells: their scalability . This scalability is consistent with the fact that neuronal trees can be described as fractal-like structures in terms of their resolution ( or complexity ) within the field in which they span rather than by their real dimensions ( see [1] ) . Based on the assumption that the principles of scalability hold true , we applied the following procedure on various dendritic structures: after all dendritic trees were centred around their somata , they were rotated manually such as to maximize the dendritic overlap . This was straightforward for cortical pyramidal cells and hippocampal granule cells where the main axis is obvious . Others , such as hippocampal pyramidal cells which will be discussed briefly further below , did not overlap since their precise branching contours depended greatly on the context and the location within the neuronal circuit . After rotating the cells into a common context , the limits of the spanning fields were measured separately for each region of the neuronal branching structure ( apical and basal dendrite for example ) . In Figure 5A such limits are shown for apical and basal dendrites of layer 2/3 , layer 4 and layer 5 pyramidal cells of the developing somatosensory cortex ( data from [29] ) . The coordinates of all nodes belonging to an individual region were scaled to the mean limits of that region within each group of cells . This resulted in size-normalized cells . Surprisingly , the scaling of the different individual region limits did not typically correlate with each other , i . e . a large apical dendrite did not necessarily mean a large basal dendrite . After scaling , the topological points belonging to one specific region could all be lumped together and a bounded density cloud was calculated ( exemplary density clouds for the cells as a whole are overlaid in Figure 5A ) . This procedure can be applied not only to different types of cells but also to different developmental stages of one particular cell type ( data from perirhinal cortical pyramidal cells [30]; see Figure 3B ) . Carrier points were then distributed randomly according to the density distributions of each region one by one and connected by the growth algorithm . The number of carrier points used was increased for each synthetic tree until it matched a target branch point number picked from the distribution observed in the real cells . In pyramidal cells , it was necessary to split the apical dendrite in a tuft region and a lower oblique dendrite region , hinting possibly to a functional or developmental requirement . The tuft was grown first and resampled to 5 µm segments to provide attaching points along the main branch for the lower part of the apical dendrite to grow on . The basal dendrite was grown separately and the resulting cells subsequently subjected to spatial jitter , soma diameter mapping and dendritic diameter tapering ( see Figure 5C and “Extended minimum spanning tree” section of Methods as well as Protocol S1 for more details ) . Using this approach , dendritic morphologies of different pyramidal cells ( Figure 6A–C: layer 2/3 , layer 4 and layer 5 , all bf = 0 . 7 ) were generated based on the spanning fields of Figure 5A . In this case , a wide range of suitable balancing factors ( bf = 0 . 4–0 . 7 ) matched the distributions of total cable length and average path lengths in the resulting synthetic morphologies leading to realistic branching statistics ( see below ) . In contrast to obtaining the density profiles for distributing random carrier points by a geometrical construction , these synthetic dendrite morphologies were obtained directly from scaled density plots from the real reconstructions . This increases the parameter space describing the generation of synthetic cells considerably . With this additional restriction , we obtain synthetic cells that are indistinguishable by eye from their real counterparts . The same approach can be used to generate pyramidal cells at different developmental stages ( Figure 6D bf = 0 . 7 , from the spanning fields of Figure 5B ) . We note that apart from the different spanning fields shown in Figure 5AB , different diameters and spatial jitter , all pyramidal cell clones were constructed according to the exact same procedure . In conclusion , Cajal's laws impose a general constraint on dendritic branching in all preparations and at all developmental stages we have investigated . Distributions of branching parameters to compare the synthetic pyramidal cells with their real counterparts are shown in Figure 7A–I . As mentioned above , because of a higher variability between pyramidal cells than for example hippocampal granule cells , branching statistics lose their informative value . Note however , that , as observed previously for insect dendrites [20] , path length distributions are similar to Sholl intersections for both real and synthetic geometries , which is a direct consequence of minimizing conduction times ( since path lengths along the tree are kept to tightly match direct Euclidean distances ) . Synthetic pyramidal cells grown with a non-optimal balancing factor on the other hand were clearly flawed , as illustrated by the example of layer 5 pyramidal neuron clones ( Figure 7K , compare bf = 0 and bf = 0 . 2 with more adequate bf = 0 . 7 ) . The electrotonic signature , developed in Figure 1 to compare electrotonic compartmentalization between neuronal trees , revealed the deficiencies of the synthetic dendrites with wrong balancing factor . When the balancing factor was too low , the electrotonic signature exhibited a distorted compartmentalization compared to an original tree ( Figure 7L , leftmost ) . This can be loosely quantified by measuring the average size of an electrotonic compartment in real morphologies compared to synthetic ones with different balancing factors ( Figure 7M ) . This quantity for a compartment size was obtained by averaging dendritic length exceeding 60% of the maximal potential deflection for current injections in all nodes one at a time . Thus , the balancing factor determines the degree of compartmentalization of the neuronal tree . This is expected , since a more stellate-like morphology associated with a higher balancing factor ( Figure 2B , right side ) should exhibit greater electrotonic segregation in its sub-trees . The local circuitry ultimately determines the context in which neuronal trees grow . There are global boundaries given by the neural tissue such as layers , topography or physical borders . However , competition for inputs between neighbouring neurons also seems to play a major role . Competition is easily implementable in the greedy growth algorithm introduced here because of the iterative nature of the algorithm . This can then be considered as a greedy extension of the growth algorithm and should be applicable in the network context . When grown under competitive conditions in which trees connect to a carrier point one after the other , the immediate consequence is spatial tiling . This can be seen in 2D for example when trees were grown from starting points on a spatial grid in a homogeneous substrate of random carrier points ( Figure 8A ) . In fact , both Cajal's material cost and his conduction time cost independently lead to this type of tiling , which does not happen in the case of random wiring ( not shown ) . Competitive dendrite growth can directly reproduce the sharp borders observed in Purkinje cell dendrites of the cerebellum . Using the geometric approach described in Figure 3 and 4 , 16 cells were grown on random carrier points distributed homogeneously in a ring-shaped area in a competitive manner . As was the case for the apical tuft of pyramidal cells , Purkinje cell dendrites required to be grown in two stages: first the thick primary dendrites and then the thinner ones covered in spines ( three sample cells are displayed in Figure 8B ) . The sharp borders of Purkinje cell dendrites could well be reproduced but whether this actually is a result of tiling in sagittal planes of the cerebellum remains to be determined experimentally . Cajal's laws can therefore explain more than just the inner branching rule: tiling between cells can emerge directly from his suggested optimization principles applied at the network level . As mentioned above , the network context also plays a major role in governing neuronal spanning fields and their density profiles . Arranging the hippocampal granule cells developed in Figure 4 onto the contours of the dentate gyrus should for example fully determine their scaling variability ( Figure 8C ) . Growing CA3 hippocampal pyramidal cells in a context-dependent manner ( here in a competitive growth process bounded by the CA3 contours of the hippocampus , from Golgi , see Ref . [31] ) might determine the variability in neuronal branching seen in the reconstructions . This is the right place to express a caveat regarding the use of branching statistics to compare dendritic structures . The branching statistics of two synthetic hippocampal pyramidal cells grown on both extremities of the CA3 region will differ entirely because of the different network context . This is the case even though these were grown using the exact same growth rule , they belong to the same cell class and they resemble their real counterparts . The idea that the network context determines a neuron's branching can be followed further: both input and output locations can serve as direct constraints for the cell morphology , as is the case when an array of photoreceptors ( Figure 8D , yellow ) in the retina connects to an array of starburst amacrine cells ( Figure 8D , green obtained as in Figure 3 ) via a set of bipolar cells . In such a case , the input-output topography of the connection determines the morphology of bipolar cells given that these grow in a competitive manner ( Figure 8D , black ) . Finally , as shown previously [20] , the growth algorithm can serve as a tool for automatic reconstruction of neuronal trees from tiled image stacks containing fluorescently labelled neurons . Figure 9A displays an example of such a tiled image stack from a small part of an insect interneuron dendrite . In order to obtain the carrier points necessary to grow the tree , the image stack is first subjected to a local threshold ( blue overlay ) which is then 3D skeletonised ( all green dots ) . The green dots are sparsened ( only larger green dots with black surround ) and a starting point is chosen ( red dot ) . Apart from the cost functions described above , additional costs can be implemented here to connect the carrier points according to the image stack information ( indicated by yellow lines between the dots ) . Connecting the carrier points results in a tree that can be further processed ( Figure 9A , bottom , green tree structure ) . The procedure can be applied competitively and in an automated or semi-automated way to recover multiple trees such as the three entangled pyramidal cells and one interneuron from a set of tiled image stacks ( Figure 9BC ) . Note that manual post-processing was required to obtain these clean reconstructions . But the simple representation of the cost function in the growth algorithm allows it to be easily extended to a state-of-the-art model-based reconstruction tool .
The spanning field in which a dendrite grows plays a major role in defining the computational and functional features of axons and dendrites . This is reflected in its importance in the process of accurately reproducing single cell morphologies . Furthermore , to replicate dendrite regions such as the apical tuft of layer 5 pyramidal neurons or the primary dendrites of Purkinje cells , a timed growth process was required in which subparts of a region were grown in a second step . This could indicate a functional constraint governing neuronal outgrowth in these cells . However , two limitations of the greedy algorithm must be considered . Firstly , the growth process does not guarantee a global optimum since it is based on an algorithm which optimizes at the local level , adding carrier points one by one . Secondly , it does not involve volumetric considerations . Both cable diameter when optimizing the amount of material used and space packing issues in conjunction with axons and dendrites of other neurons as well as with glia cells are known to play a role in determining wiring properties in the brain [3] . It is likely that some of these restrictions are responsible for the extra steps necessary in the construction of synthetic neuronal branching structures . These two possible extensions are good starting points for subsequent studies . We show that spatial tiling as observed in many dendritic structures ( Figure 8 , [32] , [33] ) is a direct consequence of Cajal's laws when applied at the network level . Indeed , network structure in general is expected to be determined by the same optimization principles , a feature which Cajal highlighted throughout his work . We have implemented this directly with the example of bipolar cells , whose carrier points were directly obtained from arrays of other existing input and output neurons rather than indirectly from its individual spanning field and its respective density profiles . By optimally arranging input and output locations , the spanning field , the major contributor to shaping the synthetic neuronal trees we have presented here , was strictly constrained by the same wiring and conduction time costs . Defining starting locations ( e . g . somata ) and synaptic target partners should therefore generally suffice to fully determine the architecture of a network . This is most likely to be a general principle and should be investigated further in future work . Based on the formalisms of optimized graphs , we have derived several new ways of representing dendritic structure and function . First , we show that graph resampling and labelling order lead to an objective representation for electrotonic compartmentalization . Simplified models which still faithfully represent the compartmentalization behaviour can be obtained with such a process . Second , taking advantage of their scalability , we derived generalized spanning fields and their density profile descriptions . These representations may be useful for comparing branching structures of different neuronal cell types . We show here how these can be useful for generating synthetic neuronal tree clones . Finally , as mentioned previously [20] after extracting carrier points directly from image stacks , the greedy algorithm is capable of a model-based automated reconstruction of neuronal trees from microscopy data . The simplicity of the algorithm and the fact that the cost factors are arbitrarily adjustable render this method an easily extendable tool . This could be crucial for combining the wide existing set of various approaches [34]–[37] in one process . For example , costs for segment orientation [37] can be integrated into the cost function directly . In summary , we find that a simple growth algorithm which optimizes total cable length and the path length from any point to the root in an iterative fashion can generate synthetic dendritic trees that are indistinguishable from their real counterparts for a wide variety of neurons . This represents a direct validation of the fundamental constraints on neuronal circuit organization described originally by Cajal . Furthermore , this approach provides a new framework for understanding dendritic tiling , which is a direct consequence of using this algorithm . The availability of these tools in a comprehensive software package ( the TREES toolbox; see Protocol S1 ) should now allow these principles to be applied to any arbitrary dendritic or axonal architecture , and permit synthetic neurons and neural networks to be generated with high precision .
The labelling of the nodes of a tree should be unique in order to , for example , compare the graphs of two trees topologically or electrotonically . In order to obtain such a unique labelling , nodes were first sorted according to their topological depth , chosen here to be the sum of the path length values of all children . Each node was then inserted in that order into a one-dimensional string , one-by-one directly behind its direct parent node . Subsequently , the resulting string of labels was mapped back onto the nodes of the tree . We refer to the Supplementary Information for more details on this subject ( Protocol S1 ) . The direct comparison of two trees along strict criteria also requires a unique distribution of node locations on the graph . We redistributed nodes on a tree structure with equal inter-nodal distances , a process we termed resampling . Each terminal branch was first lengthened by half the sampling distance . Then , starting at the root , extra nodes were positioned at integer multiples of the sampling distance along the path of the tree . All other nodes were then removed while maintaining the connectivity . During this process all segments become shorter or remain the same length; this is because a wriggly path is simplified by a straight line ( which is by definition always shorter ) . In order to preserve the total branching length and the electrotonic properties all segments were then elongated to the given sampling length ( Figure 1B , length conservation ) . Rearranging the metrics of a tree after sorting the labels leads to generation of its unique electrotonic equivalent tree . To obtain the equivalent tree , its metrics were rearranged according to a circular dendrogram where the angle towards which a segment is directed within a circle around the root corresponds to the value of the label of its target node . To obtain the electrotonic signature , the conductance matrix describing the axial conductances along the edges of the graph ( following the adjacency matrix structure ) and the membrane leak conductances ( on the diagonal of the matrix ) was simply inverted . The result is a potential matrix ( in mV ) corresponding to the resulting steady-state potential in one node when 1 nA current was injected in another node , i . e . the current transfer from any node to another . The passive axial and membrane conductances were 100 Ωcm and 2000 Ωcm2 for the sample insect dendrite sub-tree in Figure 1 and 100 Ωcm and 20000 Ωcm2 for the layer 5 pyramidal cells in Figure 7 . The branching growth was implemented as a greedy algorithm [23] as in [19] , [20] . In some cases ( Figure 5C rightmost , and all morphologies in Figures 6 , 7 and 8 ) , quadratic diameter decay was mapped on the resulting trees [19] and a soma-like increase of diameters was obtained by applying a cosine function to the diameters in close vicinity of the root ( see Protocol S1 for exact methods ) . Two-photon microscopy 3D image stacks containing neurons filled with a fluorescent dye or expressing a fluorescent protein were subject to local brightness level thresholding . After 3D skeletonization and sparsening of the resulting carrier points , these were submitted to the same greedy algorithm ( started at a user defined dendrite root location ) as used for obtaining synthetic dendrites . In the case of multiple entangled neurons as in Fig . 9BC , manual cleaning was required using the user interface provided by the TREES toolbox . See Supplementary Information for more information including the software implementing these algorithms ( Protocol S1 ) . | More than a century has passed since Ramón y Cajal presented a set of fundamental biological laws of neuronal branching . He described how the shape of the core elements of the neural circuitry – axons and dendrites – are constrained by physical parameters such as space , cytoplasmic volume , and conduction time . The existence of these laws enabled him to organize his histological observations , to formulate the neuron doctrine , and to infer directionality in signal flow in the nervous system . We show that Cajal's principles can be used computationally to generate synthetic neural circuits . These principles rigorously constrain the shape of real neuronal structures , providing direct validation of his theories . At the same time , this strategy provides us with a powerful set of tools for generating synthetic neurons , as well as a model-based approach for automated reconstructions of neuronal trees from confocal image stacks . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"neuroscience/theoretical",
"neuroscience",
"biophysics/theory",
"and",
"simulation",
"developmental",
"biology/morphogenesis",
"and",
"cell",
"biology",
"computational",
"biology/computational",
"neuroscience"
] | 2010 | One Rule to Grow Them All: A General Theory of Neuronal Branching and Its Practical Application |
Differential co-expression network analyses have recently become an important step in the investigation of cellular differentiation and dysfunctional gene-regulation in cell and tissue disease-states . The resulting networks have been analyzed to identify and understand pathways associated with disorders , or to infer molecular interactions . However , existing methods for differential co-expression network analysis are unable to distinguish between various forms of differential co-expression . To close this gap , here we define the three different kinds ( conserved , specific , and differentiated ) of differential co-expression and present a systematic framework , CSD , for differential co-expression network analysis that incorporates these interactions on an equal footing . In addition , our method includes a subsampling strategy to estimate the variance of co-expressions . Our framework is applicable to a wide variety of cases , such as the study of differential co-expression networks between healthy and disease states , before and after treatments , or between species . Applying the CSD approach to a published gene-expression data set of cerebral cortex and basal ganglia samples from healthy individuals , we find that the resulting CSD network is enriched in genes associated with cognitive function , signaling pathways involving compounds with well-known roles in the central nervous system , as well as certain neurological diseases . From the CSD analysis , we identify a set of prominent hubs of differential co-expression , whose neighborhood contains a substantial number of genes associated with glioblastoma . The resulting gene-sets identified by our CSD analysis also contain many genes that so far have not been recognized as having a role in glioblastoma , but are good candidates for further studies . CSD may thus aid in hypothesis-generation for functional disease-associations .
How can genomic information that is the same in each cell of an individual be translated into a variety of cell and tissue types ? It is clear that gene-regulatory mechanisms must play a leading role in differentiation processes . Transcription factors ( TF ) belong to the class of proteins that are able to regulate the expression of other genes . However , it is the combinatorial interactions of TFs at the promoter of a gene that determine if that gene is activated , repressed , or not regulated at all [1] . For instance during development , a tightly coordinated cascade of TFs is responsible for the activation and repression of genes that determine cell fate . The same is true for tissue specificity . The ever-increasing availability of genome-scale microarray and sequencing data has led to the development of an array of methods to investigate cells and tissues at the systems level . One class of such methods , gene co-expression analyses , has found wide use by combining microarray studies with network theory [2–9] . Investigations using a variety of network methods have found that co-expression patterns often are correlated with biologically relevant processes , such as protein-protein interactions , regulatory cascades , and biological pathways [10–16] . Because of the frequently observed relationship between co-expression and function , co-expression analyses have been used in a variety of applications . Examples include functional annotation of genes [17] , identification of pathways associated with diseases , such as Alzheimer’s [18] and autism spectrum disorder [19] , as well as inference of molecular interactions [20] . It should be noted , however , that co-expressed gene pairs do not necessarily reflect direct biological interactions: Even direct transcriptional relationships may simply be the result of accidentally matched DNA motifs without any particular function [5] . In order to facilitate the study of gene co-expression , a variety of computational tools , notably WGCNA ( [21] ) , have been made publicly available for general use . A more recent development is the study of differential co-expression networks , which seeks to identify condition-specific co-expression patterns often associated with dysfunctional regulation [22 , 23] . Many methods have been developed to generate such networks based on the implementation of different principles . In broad terms , the differential co-expression network methods can be divided into two groups . In the first group , the approaches typically generate co-expression networks that are specific to each condition studied [24–26] . Here , genes are connected by links if their co-expression score fulfills a set of statistical criteria for significance . It is a fairly straightforward matter to compare the resulting networks and subsequently , to extract interactions that are present in only one of the conditions or to identify genes subject to extensive rewiring . The second group of methods is instead focused on assigning a score for each possible gene pair , after which the score is used as input in a process to determine whether there is a significant change in co-expression between the ( possibly multiple ) conditions . These scores may be as mathematically simple as the difference between a gene-pair’s Pearson or Spearman correlation over the conditions [27] , or it may include additional steps to normalize the data [28–30] . Some of these methods determine group-wise co-expression by use of e . g . hierarchical clustering on correlation matrices [31] or decomposition of dependencies into global and group-specific components [32] . As the term “differential” suggests , the aim of both groups of methods is to identify differences in collective co-expression patterns in order to elucidate processes specifically relevant to a given condition . One example application of differential co-expression analysis is to identify target genes for treatment of a particular disorder ( e . g . a specific type of cancer ) by identifying genetic interactions potentially linked to harmful outcomes [29] . These interactions , or some of the interacting genes , may be disabled through appropriate means . However , should any of those genes also be involved in processes that are important under normal conditions , the suggested approach might run the risk of incidentally harming healthy cells and tissues . Consequently , the identification of genes that are potential disease-targets should be refined in such a way that it is possible to determine both conserved and differential co-expression in order to get a more comprehensive understanding of involved mechanisms . While the various methods differ in the measures by which they identify differential co-expression , there is also considerable variety in what sort of data they produce . Some methods only seek to identify prominent differentially co-expressed genes , without considering the genes with which they are connected [26 , 33] . Of the methods that seek to identify inter-gene relationships , some primarily focus on identifying communities or modules of genes that are collectively closely connected [28 , 30 , 34] , while others provide a more general network of differentially co-expressed genes . The latter methods can be divided into unweighted networks [25] ( also known as hard thresholding ) , in which all links are considered equal as long as they fulfill given criteria , or weighted [33] ( also known as soft thresholding ) , where links are given a numerical score quantifying their prominence . These edge weights typically represent the magnitude of change in correlation between conditions . Unweighted networks can readily be converted into weighted networks by determining a specific cut-off value for the edge weight , and uniformly setting the weight of all edges above this cut-off to 1 , while the remaining edge weights are set to 0 . It has been proposed that gene co-expression networks should follow scale-free topologies [35] , but this requirement is not universally imposed . Just as the various approaches to differential gene co-expression differ in how co-expression is determined , there are fundamentally distinct types of differential co-expression to consider . While some methods characterize differential co-expression by correlations exclusive to a given condition and others by net changes in pairwise correlations , it is important to acknowledge that these two scenarios are not entirely interchangeable . To illustrate this point , we consider the following example: Assume that a pair of genes exhibit positively correlated expression under condition α and uncorrelated expression under condition β . In this case , both classes of approaches should , in principle , be able to identify a differential co-expression . If we instead consider a pair of genes whose expression is positively correlated under condition α and negatively correlated under condition β , the two classes of methods will return different results: Methods based on net change in pairwise correlations should readily identify this gene pair as differentially co-expressed , and find the pair to be even more strongly linked than in the first example since the net change now may be larger . In contrast , methods based on determining differential co-expression by comparing unsigned co-expression networks from individual conditions [25 , 26] would not find the pair of genes to be differentially co-expressed , provided that the absolute values of the correlations are not too dissimilar . However , current “net change” methods will not be able to qualitatively distinguish between the case of positive correlation under condition α and no correlation under condition β , and the case of positive correlation under condition α and negative correlation under condition β , even though they are fundamentally different in the type of genetic correlation . In the first case , differential co-expression might suggest concerted action between the genes under condition α and independent operation under condition β . In the second case , differential co-expression suggests interactions under both conditions , but with possibly different mechanisms in play . Here , we will make a distinction between the two forms of differential co-expression in order to clearly distinguish between these two qualitatively different cases: Specific co-expression , which we will denote S , in which a gene pair is correlated under only one condition . This corresponds to the first example . Differentiated co-expression , denoted D , in which a gene pair is correlated in both tissues , but with opposing signs: In one condition , the correlation is positive and in the other condition it is negative . This corresponds to the second example . We propose a differential co-expression framework that allows the simultaneous determination and quantification of both conserved co-expressions , C , and the S- and D-types of differential co-expression patterns . Thus , a gene pair that is significantly co-expressed in one tissue may either be similarly co-expressed ( C ) , co-expressed but with an opposite sign ( D ) , or not show any significant co-expression when studied in another tissue ( S ) . In order to provide a more complete framework for differential co-expression analysis , we have developed an approach , called CSD ( “Conserved , Specific , Differentiated” ) , to categorize gene pairs according to mathematically defined scores which will allow us to construct a unified differential co-expression network from experimental data . A systematic comparison between this new method and pre-existing methods is presented under Materials and Methods . We apply the method using two types of tissues: cerebral cortex and basal ganglia from a published data set containing samples from a large number of human individuals [36] . From this network , we identify potential key sets of interactions and groups of genes which may help explain functional differences across these two tissues .
We have used the differential co-expression profiles in the GTEx V4 data set [36] , focusing on tissue groups that ( 1 ) exhibit a high degree of similarity ( as established by Pierson et al . [37] ) , and ( 2 ) for which the GTEx data set contains a sufficiently large number of samples . Our data set 1 named cortex , consists of the GTEx groups: Brain—Cortex , Brain—Anterior cingulate cortex ( BA24 ) , and Brain—Frontal cortex ( BA9 ) . Our data set 2 named basal ganglia consists of the GTEx tissue types: Brain—Caudate ( basal ganglia ) , Brain—Nucleus accumbens ( basal ganglia ) , and Brain—Putamen ( basal ganglia ) . These groupings add up to 73 data points for cortex , and 92 data points for the basal ganglia . The GTEx V4 data set contains expression data for a total 55 , 993 loci . We have restricted our analysis to protein-coding open reading frames ( as annotated in Ensembl ) , leaving a set of 18 , 453 genes ( see S4 Text ) . We calculate the pairwise gene co-expression scores ρij , k as Spearman rank-correlations for the pair of genes i and j in tissue k over all the gene expression data points N . As our analysis bases itself on identifying changes in ρij , k between conditions , we need to determine the extent of the variability of ρij , k within a condition due to confounding factors . While methods exist to evaluate the variance of computed Spearman correlations , we need to account for the fact that specific confounding factors may change the “actual” correlation within specific subpopulations of samples in a given condition . To illustrate the problem , consider a hypothetical gene-pair shown to exhibit moderate co-expression in a specific type of tissue across a whole population . Upon detailed review of the data , it turns out that this gene-pair displays very high co-expression for some particular subgroups of the population ( for instance , certain age groups , or in individuals suffering from certain diseases ) . At the same time , this pair is not showing a strong co-expression outside of these subgroups . In this hypothetical case , it would be difficult to tell if an observed difference in co-expression in another tissue and another population relates to differences between the tissues , or is due to confounding factors . On the other hand , if the genes are consistently ( but not particularly strongly ) co-expressed across all possible groupings of individuals , we can say with greater confidence that the correlation reflects genuine control related to the condition ( in this case , a tissue type ) . Consequently , we may attribute any difference of co-expression ( if similarly consistent ) in another population and another condition to the differences between the conditions . In order to determine the variance in co-expression within each given tissue , we compute the Spearman rank-correlation r i j , k l for each independent sub-sample l of size n drawn from the N data points . We use the standard error of the mean , σij , k calculated from the set of r i j , k l values , as a measure of intra-tissue co-expression variation . In order to achieve as many sub-samples as possible , increasing the chance of matching with particular confounding conditions , and accurately determine σij , k while ensuring independence between the different sub-samples , we implement the following approach for selecting sub-samples: An example of the application of this algorithm is presented in S1 Text . When implementing this algorithm , we are ensured that two data points only co-occur once in a sub-sample . Note that , it is quite beneficial to select a sub-sample size n such that n2 = N , as this greatly increases the amount of possible sub-samples that will be generated . On the other hand , a small sub-sample size n makes for a coarse Spearman correlations , i . e . to achieve three-digit accuracy , a sub-sample size of 7 is recommended . Consequently , N = 49 data points is a recommended minimum in order to determine the standard error of the correlation ρij , k within a given tissue ( or more generally , a given condition ) within reasonable accuracy . We recognize that for certain conditions ( such as those involving rare diseases or large animals which may be difficult to acquire and maintain ) this requirement might not be realistically fulfilled . In this case , subsampling may be omitted , and σij , k set to 0 for that condition ( or 1 , if subsampling isn’t possible for either condition ) . Additionally , while sub-sampling may be omitted in the event of few available data points , there would still need to be enough data points available to accurately determine the base correlation for the set . In order to determine a reasonable minimum , one should keep in mind that for two factually uncorrelated random sequences of N data points , there is still a likelihood 1/N ! that the computed Spearman correlation is 1 . As real-life gene expression data frequently involve thousands of genes , and thus millions of gene pairs , a small N can thus lead to a substantial number of false positives , no matter how stringent the required Spearman correlation . For instance , if we have a data set containing 1000 genes , and 8 expression values for each , we would expect at least 103 ( 103 − 1 ) / ( 2 ⋅ 8 ! ) ≈ 12 perfectly correlated gene pairs by pure chance in addition to any factually correlated gene pairs . While the confidence with which any perfect ( or near perfect ) correlation can be said to be biologically relevant depends on the total number of observed correlations , users should be aware of these caveats when trying to draw conclusions based on small sample sizes . Additionally , as gene expression is an inherently stochastic process , studies involving small sample sizes are susceptible to a variety of concerns with regards to noise and the resulting uncertainty in observations . Several of these are discussed in further depth in S2 Text . Based on our analysis presented in S2 Text , we hold that for comparisons on the order of thousands of genes , N = 49 data points remains a reasonably safe minimum requirement . Next , we remark upon the fact that ρij , k and its sub-sample variation σij , k are dependent: As a general rule , gene pairs with absolute correlations close to unity tend to exhibit smaller variation in co-expression between sub-samples . We present two possible explanations for this , one biological and one mathematical: If a large correlation for a pair of genes is important to cellular function , both gene products should be consistently present in nearly the same ratio in all of the sub-samples , e . g . through a process of tight gene regulation . Consequently , the observed variation in the co-expression pattern will be small . Strong correlations should therefore be more frequently associated with low variation between samples . The mathematical explanation is based on the following observation: A large sub-sample variation σij , k means that the sub-sample averages r i j , k l must follow a broad distribution . However , it is impossible for sub-sample averages to be larger than unity . Thus , when ρij , k is near unity , variation is limited in the sub-sample averages r i j , k l . Finally , an important thing to keep in mind is that as the Spearman rank correlation is less accurate for small sample sizes , a choice of larger subsample size will generally bring the subsample correlations closer to the full-sample correlation . Consequently , σij , k will generally be lower for higher subsample sizes . Because of this , it is important that the chosen sub-sample size is the same for both conditions studied in order to avoid an unbalanced contribution from one of the conditions to the denominator of Eqs 1–3 . In order to enable a systematic comparison of co-expressions , we introduce three pair-wise comparative gene co-expression scores which are computed for each pair of genes i and j in two different tissues . In general , these expressions may be applied to sets of data points from two different tissues , conditions or organisms: C i j = | ρ i j , 1 + ρ i j , 2 | σ i j , 1 2 + σ i j , 2 2 , ( 1 ) S i j = | | ρ i j , 1 | − | ρ i j , 2 | | σ i j , 1 2 + σ i j , 2 2 . ( 2 ) D i j = | ρ i j , 1 | + | ρ i j , 2 | − | ρ i j , 1 + ρ i j , 2 | σ i j , 1 2 + σ i j , 2 2 , ( 3 ) Here , Cij quantifies the extent to which co-expressions for genes i and j are conserved , i . e . similar in both tissues . Sij quantifies specific correlations: gene pairs which are strongly ( positively or negatively ) correlated in one tissue while showing no noticeable correlation in the other . Finally , Dij describes the extent to which co-expressions are differentiated: Dij is large for pairs of genes showing strong absolute correlations in both tissues or conditions , but where the nature of this correlation ( positive or negative ) changes between the two tissues . As the numerators for each of the expressions Cij , Sij and Dij are necessarily positive ( being absolute values ) , and the denominator being a positive number potentially arbitrarily close to 0 , Cij , Sij and Dij may assume any value from 0 ( included ) to infinity . However , as they follow widely different distributions , the three scores are not directly comparable within each other . In order to integrate the three types of co-expression into a common network , further steps are necessary in order to determine appropriate cut-off thresholds . We describe these in detail in the next section . Fig 1 provides a schematic visual representation of the three co-expression patterns detected by our method . Our scores are designed in such a way that they assume large values within their respective areas ( for instance , C-scores are large within the blue areas ) , while remaining small outside . Increasing the cut-off value for a given score is equivalent to shrinking the corresponding area of interest ( restricting it near the corners for C and D , and along the middle of the edges for S ) . Since these areas converge on different points as the cut-off increases , a given gene pair may not exhibit large values for more than one score . Consequently , we can choose cut-off values for each score in order to uniquely classify relevant gene pairs according to the appropriate categories . In order to generate a network for the combined gene co-expression categories , the different ranges of the scores Cij , Sij and Dij necessitate a systematic approach for combining these interaction measures . For each of the three co-expression score types , we wish to determine suitable threshold values k p C , S , D in such a way that these values correspond to the same importance level p . Thus , we will keep all gene co-expression scores C i j > k p C , and discard those below this threshold . Similarly for Sij and Dij with their respective cut-offs . However , the three different interaction scores show distributions that have noticeably different means , medians , variances and general shapes ( see S1 Fig ) . Consequently , determining whether pairs exhibit significant change or conservation based on either a common fixed-value cut-off or a given distance from the three means is incompatible with a meaningful comparison of significance across categories . Instead , we determined the importance value of a random variable X based on the likelihood of obtaining said value from the underlying distribution: If the distribution is based on M data points ( in our case , the number of different gene-pairs ) , we draw m samples si from the data set , and each si has the size L ≪ M . We determine the threshold value Xp as the average of the maximal values per sample: X p = 1 m ∑ i = 1 m max { s i } X . ( 4 ) The associated importance level is determined as p = 1/L . Thus , by choosing a common p for the gene relationship scores , we obtain a set of consistent cut-off values X p C , S , D which we use to extract separate C- , S- and D-links that are combined in a final network . It should be noted that this p is not a significance threshold , as it is determined by the distribution of the scores for a given data set , not by testing the data against a null hypothesis . Instead , its purpose is to map the scores Cij , Sij and Dij on to a common scale , to allow for meaningful comparison between them . In the consolidated network , nodes may connect to their neighbors by either C , S or D link-type . In order to distinguish between nodes predominantly involved in one type of interaction and those with multiple different types of connections , we introduce the concept of node homogeneity H: Hi=∑j∈{ C , S , D } ( kj , iki ) 2 , ( 5 ) using an expression introduced in a different context [38] . Here , kC , i , kS , i and kD , i denote node i’s number of C , S and D-type interactions , respectively , and ki is the nodes degree ( total number of connections ) . We note that in the extreme cases , H = 1 indicates a node with only one type of connections , while H = 1/3 ( the lowest possible value ) indicates a node with an even distribution of C , S and D-type connections . In order to determine enrichment of specific OMIM disease terms and KEGG pathways in our networks , we used Enrichr [39 , 40] ( http://amp . pharm . mssm . edu/Enrichr/ ) to obtain associated terms for each gene in the GTEx data set . This was a necessary step to establish accurate enrichment values , as Enrichr in itself does not provide for user-specified background gene lists . We then performed ( Bonferroni-corrected ) hypergeometric tests using NumPy ( Python ) for each of the 90 disease terms and 293 pathways listed in the Enrichr library to determine the significance of the number of associated genes in a given network . In order to perform our analysis , we developed a set of software in-house , which has been made publicly available for download ( https://github . com/andre-voigt/CSD ) . The code for computing Spearman correlations and variance was written in C++ , with the remainder ( computation of C , S , and D , estimation of cut-off , and network generation ) implemented in Python . We used Cytoscape [41] to visualize the network , and NetworkX [42] to perform network analyses . External software used for the remaining analysis is listed in S3 Text . Runtime , from expression data to the finished network may take anywhere from a few minutes to several hours for realistic data sets ( scaling approximately quadratically with both the number of genes and the number of data points . ) . As an example , for a data set consisting of 1550 genes and 100 sample points per gene , complete run time is approximately 45 minutes on an Intel Xenon X5690 CPU .
Table 1 provides a qualitative comparison between CSD and nine previously published methods for differential co-expression analysis that span a variety of method implementations . The defining characteristic of CSD is the classification of two types of differential co-expression ( S-type; the loss of co-expression in one condition , and D-type; sign change ) as well as the integration of conserved co-expression links in a composite network . Of the other listed methods , only the DCe method [33] recognizes S-type and D-type links as distinct forms of differential co-expression: it uses two different measures to determine the significance of co-expression change , one for same sign in both conditions and another when the sign changes between conditions . However , the DCe method does not distinguish between the two link-types in the final network analysis . The other listed methods either do not recognize the difference between D- and S-type links , or altogether omit D-type links from their differential co-expression analysis . Note that the CSD method explicitly includes conserved interactions in its resulting network , whereas C-type links are not included in the DCe method [33] . In order to critically validate our method’s ability to detect regulatory changes , we applied it to synthetic gene-expression data-sets with known , pre-defined regulatory interactions . As an initial test , we used a published synthetic gene data set from Zhang et . al . [43] . This has previously been used as a benchmark for relevant methods [27 , 34] . This data set consists of 20 genes , and the two conditions are defined by changing 10 of the interactions , i . e . specifying 10 differentially expressed interactions . In particular , 5 of the interactions are present only in condition 1 , and 5 ( other ) links are present only in condition 2 . We recognize all of these 10 links as S-type links within our framework , and the specified network contains no D-type links . In S1 Table , we provide detailed results of our analysis , listing the 10 top-scoring gene pairs for C-type and S-type links , as well as the top 10 S-equivalent links found by DCGL along with a classification scheme for each pair . The relevant categories in this classification , based on their connections in the reference regulatory network , are as follows: direct C ( immediate neighbors in both conditions ) , direct S ( immediate neighbors in one condition only ) , indirect C ( connected through one intermediary link in both condition ) , indirect S ( connected through intermediary links specific to one condition ) . While other interaction schemes are possible , they do not occur among the top 10 links in either test . Using our CSD-method to compute S-scores for all gene pairs , we find that the 10 direct differential interactions ( DDIs ) are assigned to 9 of the 10 top scores , with an indirect link [SWI4_SWI6 , CLB6] incorrectly assigned the 8th place . The only DDI not making it into the top 10 is the [CLB6 , MBP1_SWI6] , scoring at 11th place . In contrast , the DCe method identifies 6 of the 10 DDIs amongst its top 10 scoring S-equivalent links , and with one non-differentially co-expressed link at the 10th place . DCe does identify 2 of the 10 DDIs ( [MPB1_SWI5 , CLB6] and [MBP1_SWI6 , CLB6] ) as “switched opposites” ( equivalent to D-type in our terminology ) , but the last two selected links ( [PHO2 , CLB5] and [PHO2 , CLB6] ) are not identified as differentially co-expressed under standard parameters . Similar testing of our method focusing on the conserved links identifies 4 of the direct conserved links among the top 10 C-scores , with the remainder consisting of genes separated by a single intermediary gene . However , we note that since correlations are transitive , it is entirely within reason that gene pairs indirectly linked in the regulatory network ( but with strong co-expression along the intermediate steps ) also show strong co-expression . As this data set is unsuitable for thorough vetting of the CSD method due to the lack of D-type differential interactions , we used GeneNetWeaver [44 , 45] to generate synthetic gene-expression data from networks containing both conserved , specific and differentiated links . Starting with a general regulatory reconstruction of E . coli ( containing 1565 genes and 3758 edges ) [44 , 45] , we randomly modified 10% of the interactions ( 5% removed , 5% switched from activator to repressor , or vice versa ) . We generated 200 synthetic gene-expression data-samples for both the original and the modified network . This process was repeated 20 times ( generating new randomized networks and new synthetic expression data with 200 samples for both conditions each time ) , yielding 20 distinct sets of C , S and D-scores . To assess the quality of our method’s ranking of links , we tested true-positive and false-positive rates for C , S and D-type interactions by comparing the ranked lists of C , S , and D-type scores for the links in the known test networks . We quantify the quality by receiver operating characteristics ( ROC ) curves . We also calculated ROC curves for the DCe method on the same networks and with the same synthetic gene-expression data as input . We used the DCGL package as a benchmark , as it is , to our knowledge , the only published method that is able to distinguish between S- and D-type links ( and also demonstrates good performance in comparison to many existing methods [27] ) . Fig 2 shows the comparative ROC curves for the CSD and DCe methods . We find that on these data sets , CSD is substantially better at detecting D-type co-expression . However , despite the S-score’s success in identifying differentially expressed genes in the Zhang data set [43] , it shows weaker performance than DCe on data generated in GeneNetWeaver [44 , 45] . While we do not have any relevant method for which we can compare the performance of the C-score , we note that it’s general predictive power is higher than that of any of the other metrics . The DCe method classifies genes detected as differentially co-expressed as either S-equivalent or D-equivalent , depending on sign change in the underlying correlation , with the consequence that DCe curves do not extend across the whole range in Fig 2: Since a gene pair may only belong to one category in DCe , it is not possible to relax test requirements in such a way that one category contains all gene pairs . Notably , even under the most inclusive test requirements , the D-equivalent category can only contain on average ≈ 20% of gene pairs that show differently signed correlations between the two conditions . We observe that in general , the tested differential co-expression analysis methods do have substantial difficulties in accurately detecting individual regulatory perturbations for these types of network , and that even the best-performing measures in Fig 2 have a great deal of theoretical room for improvement in performance . This is generally the case for the existing methods [27] . A natural explanation for the difficulties in detecting individual changes lies in the fact that , empirical regulatory systems form complex networks , in which a given gene may be subjected to a multitude of regulatory impulses . Many of these input signals are shared with other genes , and regulatory cascades are common . Consequently , the loss of a regulatory interaction may not lead to a discernible change in co-expression correlations if these genes remain connected to shared regulators . On the other hand , an observed change in correlation between two genes may be the result of changes in regulatory mechanisms between intermediary genes in the regulatory network . We selected the expression data from cortex and basal ganglia from the GTEx dataset to generate a CSD network . Using an importance level of p = 10−5 on the 18453 expressed genes , we obtained a network consisting of 1814 nodes ( genes ) and 2351 edges ( Fig 3 ) . Here , transcription-factor genes are indicated by triangle-node symbol . The network contains an even mix of edge types ( 767 are C-type , 806 are S-type , and 778 are D-type ) . A link to detailed data files describing the network can be found in S4 Text . Fig 3 shows that the majority of the network is interconnected , forming a giant component consisting of 1333 ( 73 . 8% ) nodes and 2024 ( 86 . 1% ) edges . In addition to the giant component , we find 3 intermediately-sized connected components ( respectively containing: 38 nodes and 47 edges , 30 nodes and 41 edges , 13 nodes and 13 edges ) . The remaining nodes form smaller connected components: 2 components of 5 nodes and 4 edges , 8 of 4 nodes and 3 edges , 28 triplets ( all with 2 edges ) , and 137 isolated pairs . In Fig 3 , we have highlighted the names and positions of the six genes with most connections in the networks . The top-3 list of most connected nodes consists of FOXO1 ( k = 240 connections ) , ATP11C ( k = 130 connections ) , and CARHSP1 ( k = 120 connections ) , with a significant drop in connectivity to the fourth-most connected nodes ( PBX3 , k = 48 ) . A quick look at Fig 3 provides important insight concerning a key aspect of the network . It could be argued that , if conserved and differentiated interactions were functionally decoupled , and thus belonged to entirely separate parts of the genetic network , an integrated approach might not be particularly necessary , or even useful . In contrast , we find a highly interconnected network , with core regions densely interconnected by all three types of interactions . However , while the different interaction classes do not form separate networks , there is a distinct tendency for links with the same score type ( either being C , S or D ) to group together . We investigated the propensity of nodes to be connected with links of different types by calculating the homogeneity-score H ( Eq ( 5 ) ) for each node . Fig 4 ( a ) shows a box-plot of H as function of degree . Of the 1333 nodes in the giant component , 333 ( just short of half of the 701 nodes with at least 2 neighbors ) have interactions of at least two of the three different types , and 56 ( approximately 1 in 8 of the 404 nodes with at least 3 neighbors ) have interactions of all three types ( Fig 4 ( c ) ) . Interestingly , Fig 4 ( a ) suggests that highly connected genes are dominated by specific types of interactions , as shown in Table 2 . Here , of the top 5 hubs for each category , all but one of the C-hubs and one of the D-hubs have homogeneity scores over 0 . 9 . On the other end of the degree distribution , we note that nodes with very few ( less than 4 ) neighbors also tend to have somewhat more homogeneous neighborhoods than nodes with intermediate connectivity . Obviously , for nodes with only one neighbor , H = 1 , while for nodes with only two neighbors , H must be equal to either 1 or 0 . 5 ( whereas the lower bound on H for k ≥ 3 is 1/3 ) . If we classify any gene with 3 ≤ k ≤ 10 as an intermediate gene , and genes with k > 10 as a hub , we find that intermediate genes are significantly less homogeneous than hubs ( t-test , p = 3 . 8 ⋅ 10−5 ) . Looking over the combined set of intermediate and hub genes , higher degree nodes are positively correlated with increased homogeneity ( Spearman ρ = 0 . 082 , p = 0 . 085; for all genes with k ≥ 4 , ρ = 0 . 2 , p = 8 . 2 ⋅ 10−4 ) . How are the different link-types spread among the nodes ? Fig 4 ( b ) is a ternary heatmap showing a histogram of the three fractions kj , i/ki , with j ∈ {C , S , D} for each node . Consequently , entries at the corners account for all nodes with H = 1 ( see panel ( a ) ) , whereas entries at either of the sides correspond to the nodes of Fig 3 connected with only two kinds of links . For entries in the interior , the corresponding nodes are connected to all three kinds of links . For a given point on the triangle , the corresponding proportion of interactions of type X is determined by following the line parallel to the base X = 0 until reaching the base labeled X . To illustrate this , we have provided an example by marking the tile corresponding to a mix of approximately 60% C , 30% S and 10% D ( density of 19 nodes ) by a blue cross . As the densities at the corners ( representing nodes containing only one type of interaction ) are far higher than anywhere else , the color scale has been truncated at the highest non-corner value ( 0% C , 50% D , 50% S , 64 nodes ) . Whereas the Venn-diagram ( Fig 4 ( c ) ) details the number of nodes with a given mixture of links , the ternary heatmap shows how the links are mixed at the nodes . Panel ( b ) shows that the majority of the nodes connected to the three link types are dominated by C-specific ( fractions above 0 . 6 ) , and some S-specific ( near 0 . 3 ) , but only with a few D-specific interactions ( fraction near 0 . 1 ) . We evaluated the consequence of different cut-off values kp for the structure of interaction specific networks by generating separate C- , S- and D-networks for a range of importance values p ∈ [10−6 , 10−4] . For each interaction type network , we calculated their degree distribution , degree assortativity and max k-core number , and identified the top-10 most connected genes in each network . Fig 5 shows that while the C-networks exhibit greater positive degree assortativity than randomized networks with the same degree distribution , S and D networks are disassortative with respect to degree . We also find that the C-type network exhibits a higher maximum k-core value than randomized networks at the same degree distribution , while S- and D generally exhibit lower maximum k-cores ( S6 Fig ) . Both of these traits indicate that C-networks are dominated by reasonably densely interconnected sets of genes of the same type ( with highly connected genes generally connecting to other highly connecting genes , and sparsely connected genes generally connecting to sparsely connected genes ) , while the S- and D-networks follow a hub-and-spoke topology , where certain prominent genes connect to a large number of neighbors , which themselves connect to one or a few prominent nodes . We have verified that the selection of a specific cutoff in the indicated range has little effect on the topology of the resulting network , and in the case of S and D , the relative ranking of prominent hubs remains relatively constant ( S2 Table ) . Due to the comparatively small differences in connectivity of main hubs in the C-network , the changes in node rank are more pronounced , although the relative connectivity ( node degree relative to the most-connected hub ) remains stable . This allows us to select cutoffs for C , S and D to yield reasonably tractable networks , while ensuring that the specific cutoff chosen does not have a dramatic effect on key aspects of the resulting network and our presented analysis . We note that the topological differences between the C , S , and D-networks revolve around three main characteristics: degree distribution , assortativity and clustering ( defined as the proportion of common neighbors between two directly linked genes ) . In general terms , the C-network is tightly knit , with high clustering and a rather narrow degree distribution ( with less prominent hubs ) . The D-network is opposite , where most D-links connect high-degree ( k > 20 ) genes with otherwise isolated or near-isolated genes k ≤ 3 . Additionally , the clustering coefficient in the D-network is zero , and no genes that are directly connected to each other by a D-link are also connect to a common neighbor through other D-links . The S-network lies somewhere in between these two characteristics: its hubs are more prominent than those of the D-network , but less than those of the C-network , and its clustering coefficient is also somewhere in between . The differences in clustering are the result of mathematical factors—specifically , the transitivity of strong correlations . Considering three genes i , j and k , it follows from the nature of correlations that if |ρij| ≈ |ρik| ≈ 1 , then ρjk ≈ ρijρik . Because of this , if ρij and ρik remain strong and constant between conditions , then so must ρij , naturally creating “triangles” in the C-network . Assortativity is also a natural consequence of this—if gene i is strongly correlated with gene j , then j will generally tend to be correlated with i’s neighbors—therefore , if i has many neighbors , j is likely to have many neighbors as well . Similarily , should ρij and ρik simultaneously switch signs , it is mathematically not possible that ρjk also switches signs , as three genes may not all be strongly negatively correlated with each other . In fact , the D-network can be approximately characterized as a so-called bipartite network ( ignoring any potential weakening of the transitive effect over longer distances and weaker links ) . A bipartite network is defined as a network in which each node can be categorized into either of two groups , and where there are no direct links between two nodes belonging to the same group . As a direct result , a bipartite network cannot contain closed triads , and therefore has a clustering coefficient of 0 . In the case of the D-network , however , transitivity of correlations does not , by itself , adequately explain the extreme disassortativity we observe . In as much as the D-network necessarily forms a bipartite network , we noticed that the two characteristic groups roughly correspond to hubs and non-hubs . This is not a mathematical necessity—one can readily find bipartite networks in which the majority of direct connections are between hubs ( or between non-hubs ) , or with a very narrow degree distribution . We could , for instance , create a co-expression network consisting of a giant component divided into two groups of equal size , and each node connects to each of the nodes in the other groups; we could then add any number of isolated connected gene pairs . This would constitute a bipartite network consistent with a correlation network , but highly assortative . A possible explanation for the disassortativity of the D-network could reside in an argument from parsimony—that the underlying regulatory switches would happen at the individual gene level , that these are reasonably rare , and that changes to one or a few genes in a cluster would not substantially affect the relationship between the other genes in that cluster . In this case , the few perturbed genes would show D-type connections to the majority of the genes that remained constant , while the unperturbed genes would connect only to the few perturbed genes . In order to establish whether the observed network relates to possible functional aspects of the invetigated tissues , we performed GO biological process enrichment analysis using GOrilla [46 , 47] ( http://cbl-gorilla . cs . technion . ac . il/ ) on 4 networks: separately for the C , S and D-type networks generated with a draw size of 105 pairs , as well as the combined network obtained by merging the individual C , S , and D-networks . For each of the 4 networks , we found significant enrichment for a variety of biological processes ( S2 , S3 and S4 Figs ) . In all cases processes related to nervous functions are enriched , ranging from specific concepts ( e . g . regulation of neuron projection development ) to general ones ( cognition , behavior ) . Of these , GO categories for ‘anterograde trans-synaptic signaling’ is particularly prominent , showing highly significant enrichment in each of the 4 networks . It is reassuring for our method to find these GO categories as being over-represented , since we analyzed data from brain tissues . Among the remaining enriched terms , we mainly find processes related to cellular differentiation and localization , metabolism , transport and signaling . While these processes are not important for brain functions only , their enrichment in the network seems far from surprising in a co-expression network of brain tissues , given the exceptional energy requirements of the brain . While most prominent hubs in our network tend to connect to their neighbors through only one type of edge , a few genes exhibit a substantial number of connections of different types ( see Fig 4 ) . The most prominent of these is the transcription factor PBX3 . In developing macaque brains , PBX3 expression is upregulated in the basal ganglia and the cerebral cortex , suggesting a possible role in brain development [48] . However , PBX3 is mostly known as an oncogene involved in a variety of cancer types . One of these is pilocytic astrocytoma [49] ( PA ) —a form of glioma most commonly occurring in the cerebellum or areas near the brainstem ( which include the basal ganglia ) , but not in the cerebral cortex [50] , and more frequent among children and young adults [49] . The fact that our network analysis points out PBX3 as a hub with connections of different types , might hint at a molecular explanation for the differential occurrence of PA in these two tissues we analyzed . Looking at PBX3’s neighborhood , we find several other genes with similar characteristics . First , all connections this network represents are strongly positive correlations in basal ganglia . Accordingly , S-type connections correspond to weak absolute correlations in the cortex , while D-type connections correspond to strong negative correlations in the cortex . Out of 48 neighbors , 8 are suspected of influencing the development of glioma . Of these 8 genes , 6 ( SULT4A1 [51] , NDRG4 [52 , 53] , GAP43 [54 , 55] , BEX1 [56] , HINT1 [57] , LZTS1 [58] ) are believed to act as tumor suppressants , while the remaining two , PKM [59] and VIPR1 [60–62] , have been found to be overexpressed in glioma . Several of these genes also appear to play an important role in mammalian brain development and cell differentiation , where the genes VIPR1 [63] , NDRG4 [64 , 65] , BEX1 [66] and GAP43 [67] have been found to exhibit increased expression in the brain of young rats or monkeys . Using the 2016 KEGG Pathway database through Enrichr [39 , 40] ( http://amp . pharm . mssm . edu/Enrichr/ ) , we searched our network for overrepresented terms . Detailed results are provided in S3 Table . The whole network shows significant enrichment for categories including dopaminergic synapse ( S , D ) , oxytocin signaling ( S , D ) , adrenergic signaling ( D ) , glutamatergic synapse ( D ) , endocannabinoid signaling ( D ) and GABAergic synapse ( D ) . The C-network shows fewer significantly enriched pathways—the most prominent being the synaptic vesicle cycle pathway . We note that the most enriched pathways revolve around chemical compounds well known for their role on the nervous system . This is not unexpected , as our data come from two types of brain tissue . Interestingly , these pathways are not particularly well-represented in the C-network , but are ubiquitous in both the S- and D-networks . This might indicate that while these compounds play important roles across the nervous system , there might be significant regulatory differences between different types of brain tissue . In an effort to find possible causal links behind observed CSD-links , we searched the human protein interaction network ( PIN ) for connections between nodes in our network . Since CSD-links are based on co-expression analyses , it is not a given that these ( often ) indirect relationships should be reflected in direct interactions in the PIN . However , as protein-protein interactions are functionally dependent on both proteins being expressed simultaneously , we would expect these to be a potential source of C-type interactions . The PIN used for the search was compiled from three sources: the Center for Cancer System Biology’s human interactome project ( HI-II-14 ) [68] , CCSB’s literature data set ( Lit-BM-13 ) [68] , and BioGRID [69] . As the BioGRID data set is not particularly stringent when including an interaction , we decided to only include BioGRID interactions backed by at least two sources . The resulting combined PIN contains 49972 interactions for 10349 genes . 9417 of these genes are also present in the original GTEx expression data—approximately 51% of the total number of genes in the GTEx data set . Of the 1798 genes present in the combined CSD network , 1063 ( 59% ) are connected to at least one other gene in the combined PIN . This shows a moderate over-representation ( factor 1 . 16 , p < 10−6 ) of PIN genes in the CSD-network . Interestingly , 7 gene pairs are directly linked by edges in both the PIN and the CSD network ( see Table 3 ) . While this is a small section of either network , it is still a substantially larger overlap than would be expected by random chance: comparing 105 randomized versions of the CSD network ( each made by random selection of 1798 genes and 2351 gene pairs from the 18453 genes in the GTEx data ) with the PIN as the null case , we find an expectancy of edge overlap on average to be ≈ 0 . 4 , with a single case of 6 overlapping edges as the maximum observed overlap . Accordingly , the observed overlap between the actual CSD network and the PIN is approximately 18 . 7 times greater than the null hypothesis ( p < 10−5 ) . Further , we note that of these overlapping pairs , 6 are C-type edges in the CSD network ( the last pair being D-type ) . In order to investigate more indirect links , we also computed shortest paths across the PIN for each pair of nodes directly connected in the CSD-network , in order to establish whether other CSD-type connections could relate to protein interactions . We found that the genes in the differential co-expression network are more closely connected to each other than average in the PIN , with an average path distance of 3 . 95 ( against 4 . 04 for the whole network ) . While the magnitude of this effect is small , it is highly significant with p ≪ 10−3 and z = 6 . 29 ( based on the standard deviation of the mean distance of similarly sized random samples of the whole PIN ) . This suggests that the protein-protein interactions may explain certain connections in the CSD , although they are most likely not the main factor . In order to find possibly relevant mediating genes , we sorted the nodes in our PIN according to the number of shortest paths ( between genes directly connected in the CSD network ) they appeared in , with the added caveat that those paths consisted of at most 3 steps ( meaning there could be at most 2 intermediate genes in the PIN ) . The purpose of the 3-step limit being to eliminate highly indirect connections in the PIN , which are less likely to reflect an actual functional relationship . The most prominent intermediate genes in the PIN network include ESR1 , AKT1 , MDM2 , TRAF1 , UBE2I , SIRT1 and PPP1CA . Most of these genes are known to be involved in processes which should be relevant to the differentiation and function of neural tissue , such as regulation of gene expression ( ESR1 , AKT1 , UBE2I , SIRT1 , PPP1CA ) and metabolic/catabolic processes ( AKT1 , MDM2 , UBE2I , SIRT1 , SKP2 , PPP1CA ) . In more specific detail , ESR1 and TRAF1 are both involved in regulation of NF-kB signaling—ESR1 as an inhibitor and TRAF1 as an activator . NF-kB is known to be involved in synaptic plasticity , learning , and memory , and may be activated by synaptic transmission . Promoter hypermethylation at ESR1 [70] , expression of TRAF1 [71] and mutations in NF-kB [72] are all known to be associated with the emergence of glioma . AKT1 is known to interact with forkhead box transcription factors [73] ( which include FOXO1 , the most highly connected node in the differential co-expression network ) in order to regulate cell growth and apoptosis . As FOXO1 connects to several of the cancer-associated genes adjacent to PBX3 ( though not to PBX3 directly ) , the relative prominence of both FOXO1 and AKT1 might reflect a potential tumor-inhibiting effect in combination with PBX3 and its neighbors . As we were able to identify a number of key neurological process amongst the genes present in our networks , we sought to investigate if there could be any links between differential co-expression and inheritable disease . Using the extended OMIM disease association data set , we found no significant enrichment for disease-associated genes in general in the combined network ( p = 0 . 417 ) or in the C-only or D-only networks ( p = 0 . 175 and p = 0 . 649 , respectively ) . However , we did find an over-representation ( by a factor of 1 . 35 ) of disease-associated genes amongst genes in the S-network ( non-corrected p = 0 . 00458 ) . Using Enrichr [39 , 40] ( http://amp . pharm . mssm . edu/Enrichr/ ) to search the C , S , D- and combined networks for specific OMIM disease associations , we find substantial enrichment for one of two disease families , depending on the kind of network . The S-network shows an over-representation of genes associated to epilepsy , with 8 genes in the network , while only 1 . 7 genes would be expected by chance ( 4 . 65-fold enrichment , p < 0 . 4 ⋅ 10−3 ) . The D-network shows enrichment for ataxia , with 6 genes ( expected number 1 . 1 , 5 . 4-fold enrichment , p < 8 ⋅ 10−3 ) , and more specifically , spinocerebellar ataxia , with 5 genes ( expected number , 0 . 62 , 8-fold enrichment , p < 2 ⋅ 10−3 ) . While both terms are rather broad and may refer to any of a variety of diseases with different underlying mechanisms , they both involve defects in motor functions , which are controlled by basal ganglia and cerebellum . Noting that three of the four dominant hubs exhibit protein interactions with a number of glioma-related genes , as well as the presence of several glioma-related genes in PBX3’s neighborhood , we mapped out their immediate network ( see Fig 6 ) . Furthermore , we performed an exhaustive literature search to identify whether any of FOXO1 or CARHSP1’s immediate neighbors also exhibited particular expression patterns related to glioma . In fact , in the combined neighborhoods of FOXO1 , CARHSP1 and PBX3 , we find 104 such genes ( out of a total of 340 in said neighborhoods ) ( see Table 4 for a detailed listing ) . For most of these genes ( 59 ) , increased expression is associated with beneficial outcomes , while 45 genes have their activity linked to increased proliferation , invasiveness and general mortality . The hubs themselves are all associated with aggressive forms of glioma . As previously mentioned , PBX3 is known to be upregulated in PA [49] , while increased CARHSP1 expression is linked to necrosis and microvascular proliferation ( MVP ) [74] . On the other hand , FOXO1 is known to prevent cell proliferation in glioblastoma [75] . We also note that the gene-glioma associations presented come from a variety of previously performed studies , and that there is no guarantee that the literature contains an exhaustive list of genes involved in glioma . It is therefore quite possible that there are genes important to glioma development whose role has not yet been discovered , and consequently , would not have been identified here . The substantial presence of known glioma-associated genes in the neighborhood of FOXO1 , PBX3 and CARHSP1 may indicate the additional presence of genes with currently unidentified roles in glioma . We therefore present the exhaustive neighborhoods of FOXO1 , PBX3 and CARHSP in S4 Table Text as candidate genes for further study .
In this paper , we describe a new method for identifying differential co-expression relationships between genes when comparing two tissues . In contrast to previous methods , our method allows the detection of genes that play critical roles in context-specific function , based on similarities and differences in co-expression patterns . We demonstrate the power of our new method by analyzing the network of cortex and basal ganglia tissues , which is revealed to be associated with a variety of important aspects of brain function . In particular , we find substantial enrichment of ( 1 ) GO terms such as anterograde synaptic signaling , cognition , and neural development , ( 2 ) hereditary links to the neurological diseases ataxia and epilepsy , and ( 3 ) genes associated with pathways involving compounds important to brain function , such as adrenaline , dopamine and oxytocin . Furthermore , we find indications for the hub PBX3 to be involved in the occurrence of PA—which occasionally occurs in basal ganglia but not in the cortex . In addition , we suggest that the general preponderance of GO terms with clear relevance to brain development and function indicates that the resulting network represents genuine and meaningful relationships between the genes present in the network . While the gene expression data used in this study came from the same source , this is not a requirement for the method to be viable . Since the networks are based on the non-parametric Spearman rank-correlation ( which relies only on the relative rank of each data point within its set ) calculated within each of the compared data sets , it is not necessary for the expression values in the different sets to be normalized against each other . In fact , one could compare a tissue with log-scale expression values ( e . g . coming from microarrays ) against one where the expression values follow a linear scale ( e . g . RNA-Seq data ) , without any impact on the resulting network . It should be noted , however , that the networks obtained by this method do not correspond to protein-protein interaction networks or even gene regulatory networks , and that the presence of a link between two genes in the differential co-expression network does not necessarily reflect any direct biological interaction between the two . In fact , a link is only evidence of a coinciding pattern: two co-expressed genes may both be regulated by a common transcription factor , or may be similarly affected by outside factors ( for instance , nutrient availability ) . We note that triple-type nodes ( involved in all three types of interactions ) are dominated by C-type interactions and a very small share of D-type interactions . We also note that the leading D-type hubs have far more connections than those of the other types . This may suggest that the D-type regulatory change between tissues demonstrates a much more concentrated effect: even if the underlying changes are focused near only a few key genes , a disproportionately large amount of interacting genes could be affected . The key topological difference between the C-type network on one hand ( highly assortative and with a substantial densely connected core ) and the S- and D-type networks on the other ( with a few dominant hubs , especially in the case of D ) also indicate a possible difference regarding the regulatory mechanisms involved . Hence , we speculate that a tightly co-regulated cluster of genes might involve more redundant ( and thereby robust ) regulatory mechanisms and therefore be less likely to change . In contrast , genes with more centralized neighborhoods may be more likely to see large changes in co-expression patterns due to perturbations at the individual gene levels . An alternative hypothesis is that the strong prominence of hubs in the D-network comes as a result of regulatory changes mostly involving a few genes within large co-expressed clusters , whereby the few perturbed genes would form D-type links with the remainder of said clusters . We take particular note of a set of gene clusters , centered around the transcription factors FOXO1 , CARHSP1 and PBX3 . These consist of multiple genes believed to be of major importance to both neural development and the emergence of glioma . While it is known that defects in genes controlling growth and differentiation is a common factor in cancers , to the best of the authors’ knowledge , no association between these specific genes has previously been determined . However , it is difficult to present conclusions about the underlying cause of the observed co-expression patterns as certain . While the FOXO1/PBX3/CARHSP1-centered gene clusters suggest a functional link between several glioma-associated genes , it does not , for instance , automatically follow that misregulation of ( or by ) PBX3 is the key driver in glioma development . While it is hard to determine a definite cause behind these connections at the gene level , a comparative study between the CSD network and the PIN offers one possible explanation . We find that in the PIN , both PBX3 and CARHSP1 are indirectly connected to each other ( as well as several of their other neighbors in the CSD-network ) through the intermediary of TRAF1 , whose overexpression is also associated with the emergence of glioma [71] . We also find similar intermediary protein interactions through ESR1 , AKT1 and SIRT1 , whose activity are also associated with glioma [70 , 168 , 169] . Additionally , AKT1 is known to interact with FOXO1 ( the most prominent hub in the CSD network ) to inhibit apoptosis [73] . FOXO1 and AKT1 , along with MDM2 ( another common intermediary gene in the PIN ) have previously been identified in differential co-expression studies of glioblastoma [170] . The prominence of glioma-related cells in these clusters is somewhat unexpected , as our comparison is not between cancerous and non-cancerous data sets , but rather of two ( nominally healthy ) different parts of the brain . However , we note a substantial overlap between glioma-associated genes and genes particularly expressed in developing ( embryonic and juvenile ) brains . Additionally , GO enrichment tests using the Gene Ontology Consortium database ( www . geneontology . org ) [171] return an 1 . 8-fold enrichment ( Bonferroni-corrected p = 2 . 1 ⋅ 10−2 ) for the term “nervous system development” amongst the 340 genes in the neighborhood of FOXO1/CARHSP1/PBX3 , and a 2 . 7-fold enrichment ( corrected p = 1 . 2 ⋅ 10−3 ) for the same term among the 104 genes for which we found associations with glioma . The latter 104-gene set also shows significant enrichment for more specific subterms of “nervous system development” , including “neuron differentiation” ( 3 . 5-fold , p = 4 . 9 ⋅ 10−3 ) , “neuron projection morphogenesis” ( 5 . 4-fold , p = 2 . 3 ⋅ 10−2 ) and “axonogenesis” ( 6 . 26-fold , p = 1 . 45 ⋅ 10−2 ) . The observed connection between these genes may therefore reflect a role in the differentiation of stem cells into specific types of brain tissue . Hence , it is plausible that perturbations in these differentiating mechanisms result in differentiation of brain cells into cancerous tissue , which would explain why so many of these genes emerge in studies involving gene expression in glioma . The scope of the method for differential co-expression network analysis presented in this paper is not restricted to only comparing two different tissues within a given organism . In fact , it may be used to compare any two sets of gene expression data for which a comparison might be reasonable: the only criterion is the existence of a viable one-to-one match between the genes in each data set . Possible applications of our method include comparing gene expressions between healthy and sick individuals , comparing samples from experiments with before/after treatments , comparing organisms subjected to different external environments and comparing closely related species with known orthologs . | With the ever increasing availability of large sets of gene expression data , much effort has been directed towards studying shared expression patterns between different genes . We have developed a general method for studying the variation of gene co-expression between two different conditions , which allows for a more detailed description and classification of interactions than previous methods . Applying our method to compare data from two different parts of the brain ( the cortex and the basal ganglia ) , we find that it identifies genes known to be involved in key brain functions . Our analysis also identifies connections between a variety of genes previously known to be involved in the progression of glioma . Our method can also be applied in studies comparing between healthy and disease states , treatment and controls , among others . | [
"Abstract",
"Introduction",
"Materials",
"and",
"methods",
"Results",
"Discussion"
] | [
"medicine",
"and",
"health",
"sciences",
"genetic",
"networks",
"neural",
"networks",
"gene",
"regulation",
"protein",
"interaction",
"networks",
"cancers",
"and",
"neoplasms",
"brain",
"neuroscience",
"oncology",
"neurological",
"tumors",
"network",
"analysis",
"computer",
"and",
"information",
"sciences",
"basal",
"ganglia",
"gene",
"expression",
"proteomics",
"biochemistry",
"anatomy",
"gene",
"identification",
"and",
"analysis",
"neurology",
"genetics",
"biology",
"and",
"life",
"sciences",
"glioma"
] | 2017 | A composite network of conserved and tissue specific gene interactions reveals possible genetic interactions in glioma |
We investigated the growth properties and virulence in mice of three Zika virus ( ZIKV ) strains of Asian/American lineage , PRVABC59 , ZIKV/Hu/Chiba/S36/2016 ( ChibaS36 ) , and ZIKV/Hu/NIID123/2016 ( NIID123 ) , belonging to the three distinct subtypes of this lineage . The American-subtype strain , PRVABC59 , showed the highest growth potential in vitro , whereas the Southeast Asian-subtype strain , NIID123 , showed the lowest proliferative capacity . Moreover , PRVABC59- and NIID123-infected mice showed the highest and lowest viremia levels and infectious virus levels in the testis , respectively , and the rate of damaged testis in PRVABC59-infected mice was higher than in mice infected with the other two strains . Lastly , ZIKV NS1 antigen was detected in the damaged testes of mice infected with PRVABC59 and the Pacific-subtype strain , ChibaS36 , at 2 weeks post-inoculation and in the epididymides of PRVABC59-infected mice at 6 weeks post-inoculation . Our results indicate that PRVABC59 and ChibaS36 exhibit increased abilities to grow in vitro and in vivo and to induce testis damage in mice .
Zika virus ( ZIKV; genus: Flavivirus; family: Flaviviridae ) was first isolated from a sentinel rhesus monkey in the Zika forest of Uganda in 1947 [1] . ZIKV infection in humans was first identified in Uganda and the United Republic of Tanzania in 1952 , and human ZIKV infections ( ZIKV disease , ZVD ) have been sporadically detected in Africa and Asia for > 50 years since the initial isolation [2] . The first large outbreak of ZIKV infection in humans was identified in Yap Island in the Federal State of Micronesia in 2007 and a ZIKV-infection outbreak was also confirmed in French Polynesia in the South Pacific in 2013 , with an estimated number of 30 , 000 ZVD cases [3–5] . In 2014–2015 , ZIKV-infection epidemics spread to other Pacific regions and to the Americas . In 2016 , patients with ZVD were reported in several countries in Southeast Asia , including Singapore , Thailand , Vietnam , and the Philippines . ZIKV is transmitted to humans mainly through the bite of Aedes mosquitoes and non-vector transmission of ZIKV has also been reported to occur through blood transfusion , transplantation , and sexual intercourse [6–8] . The clinical symptoms caused by ZIKV infection are generally self-limiting and include fever , rash , headache , joint and muscle pain , and conjunctivitis . Approximately 60–70% and 90% of symptomatic ZIKV-infected patients develop fever and rash , respectively . When pregnant women are infected with ZIKV , the fetus can be infected with ZIKV through the placenta , which can cause congenital neurological malformations with the following symptoms: microcephaly , sensorineural abnormalities , cerebral calcification , and abortion [9 , 10] . ZIKV infection also causes a severe neurological complication , Guillain–Barré syndrome , and ZIKV has demonstrated the ability to infect diverse cell types including neuronal cells [9 , 11 , 12] . The ZIKV genome is comprised of a single-stranded , positive-sense RNA that encodes three structural proteins ( C , prM , and E ) and seven non-structural proteins ( NS1 , NS2A , NS2B , NS3 , NS4A , NS4B , and NS5 ) in one open reading frame [13] . ZIKV is classified into two lineages , African and Asian/American; the recent epidemics associated with severe neurological and congenital abnormalities in Pacific regions and the Americas were caused by the spread of Asian/American-lineage ZIKV from Southeast Asia [14 , 15] . Phylogenetic analyses performed using complete ZIKV genomes indicate that ZIKV strains in the Asian/American lineage can be divided into three subtypes , American , Pacific , and Southeast Asian , which present several differences in their amino acid sequences [16 , 17] . Studies conducted using a mouse model showed that American-subtype ZIKV strains induce more severe neurological disorders and marked immune responses in mice compared with Southeast Asian-subtype strains [18 , 19] . These findings raise the possibility that recent genetic changes in the ZIKV genome altered the properties of the virus , such as virulence and tissue tropism , and also contributed to the spread of the ZIKV endemic area and the increase in the cases of congenital ZIKV infections in Pacific regions and the Americas [17] . Accordingly , mutations in prM and NS1 were reported to be involved in the differences between Asian/American-lineage strains in terms of pathogenicity in mice , viral protein antigenemia , and interferon induction in host cells [18 , 20–24] . However , only a limited number of Asia/American-lineage ZIKV strains were previously used for evaluating virulence; thus , further analyses must be conducted using additional ZIKV isolates to elucidate the relationship between genetic variation and pathogenicity among the Asian/American-lineage strains [25] . In this study , we examined the in vitro and in vivo growth of three ZIKV strains that belong to the three distinct subtypes in the Asian/American lineage . Recent studies have demonstrated that Asian/American lineage ZIKV infection in mice is not lethal , but causes testis damage; this indicates that testis damage in ZIKV-infected mice can be used as an index for assessing ZIKV pathogenicity [26–33] . Therefore , in this study we pathologically evaluated the male reproductive organs of mice infected with each of the three strains to determine the differences in pathogenicity among these ZIKV strains .
Mouse experiments were performed in biosafety level 2 animal facilities , in accordance with the “Guidelines for Animal Experiments Performed at the National Institute of Infectious Diseases ( NIID ) , ” under approval ( no . 116067 ) from the Animal Welfare and Animal Care Committee of the NIID , Japan . All efforts were made to minimize any potential pain and distress . Mice infected with ZIKV were observed daily for adverse reactions and signs of diseases . For collection of organ samples , mice were euthanized by using isoflurane . The ZIKV strains used in this study were previously published and we did not obtain samples from patients specifically for this study [16 , 34 , 35] . We used four ZIKV strains: MR766-NIID ( MR766; accession no . LC002520 ) , ZIKV/Hu/Chiba/S36/2016 ( ChibaS36; accession no . LC191864 ) [35] , PRVABC59 ( accession no . KU501215 ) , and ZIKV/Hu/NIID123/2016 ( NIID123; accession no . LC219720 ) [16] ( Table 1 ) . ChibaS36 and NIID123 were originally isolated from , respectively , a patient infected with ZIKV in Fiji in 2016 and a patient infected with ZIKV in Vietnam in 2016 [16 , 34 , 35] . PRVABC59 was kindly provided by Dr . Beth Bell of the US CDC . MR766 was maintained at the NIID , Japan; this strain was used as a positive control when evaluating the growth ability of Asian/American-lineage ZIKV in mice [34] . These viruses were propagated in Vero cells ( strain 9013 ) . Vero cells and mosquito-derived C6/36 cells were cultured at 37 °C and 28 °C , respectively , under 5% CO2 in Eagle’s minimum essential medium ( MEM ) supplemented with 10% heat-inactivated fetal bovine serum ( FBS ) and 100 U/mL penicillin–streptomycin ( MEM-10FBS ) . The infectious titer of the viruses was determined using the plaque assay , as previously described [36] . Growth kinetics was analyzed in Vero cells and C6/36 cells as previously described [34] . Briefly , Vero and C6/36 cells were plated in 6-well culture plates ( 3 × 105 cells/well ) and infected with ZIKV at a multiplicity of infection of 0 . 01 plaque-forming units ( PFU ) per cell . Small aliquots of the culture medium were collected periodically and the titer of the infectious ZIKV in each aliquot was determined using a plaque assay performed on Vero cells . Growth curves were statistically compared using BellCurve for Excel ( Social Survey Research Information , Tokyo , Japan ) employing a two-way ANOVA test . To construct a phylogenetic tree , the nucleotide sequences of 33 ZIKV strains were aligned and analyzed using the maximum likelihood method with 1 , 000 bootstrap replicates using the MEGA7 program [37] . Complete amino acid sequences of the ZIKV strains were aligned using GENETYX gene-analysis software ( Genetyx Corp . , Tokyo , Japan ) . Interferon-α/β receptor-1 gene-knockout ( IFNAR1-KO ) C57BL/6 mice were produced , bred , and maintained in a specific-pathogen-free environment as previously described [34 , 38 , 39] . Mice ( 8–16 weeks old ) were inoculated with 1 × 104 PFU of each virus in culture medium diluted with MEM supplemented with 2% FBS ( MEM-2FBS ) through the subcutaneous route in the footpad . Tissue and fluid samples ( serum , brain , spinal cord , kidney , spleen , testis , epididymis , sperm , and epididymal fluid ) were collected from the ZIKV-infected mice and the infectious virus and viral RNA levels in the samples were measured . Tissue weights were also determined prior to homogenate preparation . To prepare sperm and epididymal fluid samples , cauda epididymis was placed in a microtube containing 100 μL of phosphate-buffered saline ( PBS ) and incised 4–5 times with scissors to allow the sperm to swim out and disperse , after which the cauda epididymis was removed and the liquid phase was centrifuged for 5 min at 10 , 000 × g . The supernatant was recovered as the epididymal fluid and the precipitate was resuspended in 100 μL of PBS and used as the sperm sample . The collected tissues ( 10–200 mg ) were homogenized in 500 μL of MEM-2FBS and then used for further analyses . We determined the 50% tissue-culture infective dose ( TCID50 ) for each organ sample as previously described [30 , 40] . Briefly , to pre-amplify infectious ZIKV in the organ samples using C6/36 cells , the organ samples were serially diluted ( 1:10–1:108 ) with MEM-2FBS and the C6/36 cells cultured in 96-well plates were then inoculated with each sample and incubated for 5 days at 28 °C . The pre-amplification step was added to improve the detection of low-level infectious viruses . Subsequently , 25-μL aliquots of the culture supernatants were transferred to Vero cells cultured in 96-well plates and incubated for 5 days at 37 °C . Lastly , the cells were fixed in 10% formaldehyde solution and stained with Methylene Blue solution to visualize the cytopathic effect induced by the ZIKV infection . Viral titers were statistically compared using either BellCurve for Excel employing the Mann-Whitney U test or SPSS ( IBM , Chicago , IL , USA ) employing the repeated-measures ANOVA test . Total RNA was extracted from each organ sample using a High Pure Viral RNA Purification Kit ( Roche Diagnostics , Indianapolis , IN , USA ) . ZIKV RNA was quantified by performing quantitative real-time RT-PCR with Fast Virus One-Step Master Mix ( Thermo Fisher Scientific , Waltham , MA , USA ) . The ZIKV genome was amplified using the following primers: ZIKV 1086 ( 5′-CCGCTGCCCAACACAAG-3′ ) , ZIKV 1162c ( 5′-CCACTAACGTTCTTTTGCAGACAT-3′ ) , and Probe ZIKV 1107-FAM ( 5′FAM-AGCCTACCTTGACAAGCAGTCAGACACTCAA-TAMRA3′ ) [14] . Genome copy numbers were statistically compared using BellCurve for Excel employing the Mann-Whitney U test . Testes and epididymides were fixed in 10% phosphate-buffered formalin , embedded in paraffin , sectioned , and stained with hematoxylin and eosin . Immunohistochemistry was performed using an anti-ZIKV NS1 antibody ( C01886G , Meridian Bioscience , Cincinnati , OH , USA ) as the primary antibody . Specific antigen–antibody reactions were visualized by 3 , 3-diaminobenzidine tetrahydrochloride staining using a DAKO LSAB2 system ( DAKO Cytomation , Glostrup , Denmark ) .
The phylogenetic analysis results indicated that the three ZIKV strains , PRVABC59 , ChibaS36 , and NIID123 , belong to distinct subtypes in the Asian/American lineage: PRVABC59 , American subtype; ChibaS36 , Pacific subtype; and NIID123 , Southeast Asian subtype ( Fig 1 ) . The plaques formed on Vero cells by NIID123 were clearly smaller than those formed by each of the other three strains ( Fig 2A ) . NIID123 also replicated more slowly than the other three ZIKV strains in Vero cells ( Fig 2B ) . The growth kinetics of NIID123 and ChibaS36 in C6/36 cells revealed a slower replication rate than that of MR766 and PRVABC59 ( Fig 2C ) . No statistically significant differences in viremia levels were observed among the four ZIKV-infected mouse groups at 6 days post-inoculation ( Fig 3A ) . However , no infectious virus was detected in the serum of any mice infected with NIID123 . Viral RNA was detected in all mice infected with each strain; however , the viral RNA level in the NIID123-infected mice was the lowest among the four groups ( Fig 3B ) . Infectious virus was detected in the brain and spinal cord of all mice infected with MR766 , but not in any of the mice infected with the Asian/American-lineage strains ( Fig 3A ) . No infectious virus was detected in the liver and kidney of most of the mice infected with each of the four ZIKV strains . Infectious virus was detected in the spleen in 2/3 of ChibaS36-infected mice , but not in mice infected with PRVABC59 or NIID123 . Viral RNA was detected in all tissues of mice infected with MR766 , PRVABC59 , or ChibaS36 ( Fig 3B ) . Viral RNA was not detected in the brain or spinal cord of 50% of the mice infected with NIID123 . Furthermore , the viral RNA level in the tissues of mice infected with NIID123 was the lowest among the four groups . Next , we analyzed the three Asian/American-lineage ZIKV in further detail . The time course of change in viremia levels in mice infected with the three strains was also examined ( Fig 4 ) . Mice infected with ChibaS36 showed significantly lower and higher viremia levels compared with mice infected with PRVABC59 and NIID123 , respectively . All inoculated mice survived to the final collection point ( 8 or 9 days post-infection ) . All inoculated mice had survived at 2 weeks post-inoculation . No infectious virus was detected in the serum in any of the mice infected with each Asian/American-lineage ZIKV strain ( Fig 5A ) . Infectious virus was detected at similar levels in the epididymal fluid and epididymal cells including sperm ( sperm/epididymal cells ) samples from all mice infected with PRVABC59 or ChibaS36 , whereas infectious ZIKV was not detected in the samples from NIID123-infected mice . Infectious virus was detected in the testes of all PRVABC59-infected mice , but not in the testes of ChibaS36- or NIID123-infected mice . Lastly , infectious virus was detected in the epididymis of some of the mice infected with each strain , but no significant difference was observed between the three groups . Viral RNA was detected in all specimens collected from PRVABC59- and ChibaS36-infected mice at 2 weeks post-inoculation ( Fig 5B ) . Serum viral RNA levels were significantly higher in PRVABC59-infected mice than in ChibaS36-infected mice . Viral RNA was also detected in the serum of NIID123-infected mice , but the levels were significantly lower than those in the serum of PRVABC59- and ChibaS36-infected mice . Viral RNA levels in the testis were significantly higher in PRVABC59-infected mice than in ChibaS36-infected mice , whereas viral RNA levels in the epididymis , epididymal fluid , and sperm/epididymal cell samples from PRVABC59-infected mice were similar to those in the corresponding samples from ChibaS36-infected mice . All inoculated mice survived at 6 weeks post-inoculation . Infectious virus was not detected in any of the serum samples and most of the genital samples collected from mice infected , except in the sperm/epididymal cell specimen from one PRVABC59-infected mouse ( Fig 5C ) . Viral RNA was identified in the serum , epididymal fluid , sperm/epididymal cell , testis , and epididymis specimens of PRVABC59-infected mice and ChibaS36-infected mice ( Fig 5D ) . The levels of viral RNA in the serum , epididymal fluid , and sperm/epididymal cell specimens of PRVABC59-infected mice were higher than the levels in the specimens of ChibaS36-infected mice . No viral RNA was detected in any of the specimens of NIID123-infected mice . At 2 weeks post-inoculation , no weight loss was observed in the testis of mice infected with each ZIKV strain ( Fig 6A and S2 Table ) . However , at 6 weeks post-inoculation , testis weight was decreased in 5/6 ( 83 . 3% ) PRVABC59-infected mice and 1/6 ( 16 . 7% ) ChibaS36-infected mice; the two groups of mice showed significant differences in the frequencies of testis damage ( P = 0 . 04 , Fisher’s exact test ) , but not in testis weight ( Fig 6B and S2 Table ) . Testis weight loss was not observed in NIID123-infected mice and the frequency of testis damage was also significantly lower than that in PRVABC59-infected mice ( P = 0 . 0076 , Fisher’s exact test ) . In all three groups of ZIKV-infected mice , the epididymides were smaller at 6 weeks post-inoculation than at 2 weeks post-inoculation , but epididymis weight did not differ among the groups at the same time points ( Fig 6C , 6D and S2 Table ) . Moreover , sperm mass was not obtained from the cauda epididymis of PRVABC59- and ChibaS36-infected mice with damaged testes . At 2 weeks post-inoculation , germ cells were necrotic and inflammatory cells , including neutrophils , had infiltrated into the testicular interstitium in PRVABC59- and ChibaS36-infected mice ( Fig 7A , 7C , 7G , and 7I ) . At 6 weeks post-infection , the testes of PRVABC59-infected mice were atrophic and most germ cells were not observed in the testis ( Fig 8A ) . The ZIKV NS1 antigen was detected in the seminiferous tubules and epididymal fluid of PRVABC59- and ChibaS36-infected mice at 2 weeks post-inoculation ( Fig 7B , 7D , 7F , 7H , 7J and 7L ) , whereas the antigen was only detected in the epididymides of PRVABC59-infected mice at 6 weeks post-inoculation ( Fig 8D ) .
Previously , we reported that Asian/American-lineage ZIKV strains can be classified into three subtypes based on amino acid sequences: the Southeast Asian subtype , Pacific subtype , and American subtype [16] . In the present study we showed that ZIKV strains of these three subtypes exhibited distinct growth properties in vitro and in vivo . ZIKV NIID123 , which belongs to the Southeast Asian subtype , showed lower growth capacity in vitro and in vivo and exerted a weaker injurious effect on the testis and epididymis of mice compared with the other two strains . ZIKV PRVABC59 showed the highest growth capacity both in vitro and in most of the reproductive organs examined in mice and also induced maximal damage in the testis and epididymis of the three strains . These results raise the possibility that ZIKV have acquired elevated proliferative capacity and pathogenicity during the process of the virus spreading from Southeast Asia to the Americas through the Pacific islands . The PRVABC59 and ChibaS36 strains clearly exhibited higher growth ability than NIID123 strain in Vero cells ( Fig 2B ) . The passage numbers of the three strains in Vero cells were ≤ 7; however , NIID123 was first passaged in C6/36 cells ( Table 1 ) . Therefore , we cannot exclude the effect of NIID123 passage history on its growth in Vero cells and mice . However , the growth rate of NIID123 in C6/36 cells was also clearly lower than PRVABC59 , indicating that NIID123 has not adapted to C6/36 cells ( Fig 2C ) . Our results demonstrated that infection with each of the three Asian/American-subtype strains used in this study was not lethal to IFNAR1-KO mice , whereas we previously demonstrated that infection with the African-lineage MR766 strain was lethal when the mice were inoculated with a lower infectious titer ( 1 × 102 PFU ) of strain MR766 [34] . In MR766-infected mice , high levels of infectious virus and viral RNA were detected in the central nervous system at 6 days post-infection . In contrast , no infectious virus was detected in the central nervous system in mice infected with any of the Asian/American-lineage strains ( Fig 3 ) . These results support the hypothesis that the African-lineage strain causes more acute infection than the other ZIKV strains [41 , 42] . However , this does not exclude the possibility that MR766 was adapted to the mouse central nervous system because of the repeated past passaging of this strain in mouse brains for virus isolation and maintenance . In this study we examined the viral loads in the brain , spinal cord , liver , kidney , and spleen in ZIKV-infected mice at only one time point ( 6 days post-infection ) ; therefore , further analysis may be needed to evaluate the detailed kinetics of viral load in the organs . At 2 weeks post-infection , the amounts of infectious particles and viral RNA in the epididymis , sperm/epididymal cells , and epididymal fluid were almost equal in PRVABC59- and ChibaS36-infected mice , but were clearly lower in NIID123-infected mice ( Fig 5A and 5B ) . Conversely , although no infectious particles were detected at 6 weeks post-infection , higher rates and titers of the viral genome were detected in sperm/epididymal cells and epididymal fluid from PRVABC59-infected mice than from ChibaS36-infected mice ( Fig 5C and 5D ) . These findings suggest that the American-subtype strain might be maintained for a longer period than the Pacific-subtype strain in infected mice and that the Southeast Asian-subtype strain might possess the lowest proliferative potential or might be rapidly excluded from the infected mice . At 2 weeks post-infection , the infectious virus titer and viral RNA levels in the testis were markedly high in PRVABC59-infected mice , but no infectious virus was detected in ChibaS36- and NIID123-infected mice ( Fig 5A and 5B ) . These results raise the possibility that PRVABC59 not only induces higher levels of viremia , but also exhibits a higher ability to proliferate in the testis compared with the other two strains . At 6 weeks post-infection , viral RNA was detected in the testis in 5/6 PRVABC59-infected mice and 1/6 ChibaS36-infected mice and all the testes in which viral genomes were detected were damaged ( Figs 5 , 6 and 8 ) . These data provide compelling evidence that persistent infection of testis with American- or Pacific-subtype ZIKV might represent the cause of testis damage in ZIKV-infected mice . The observed phenotypes of the damaged testes in our experiments resembled those previously described [26 , 28 , 29] . The seminiferous tubule in the damaged testes showed atrophy and scarring due to germ-cell loss and inflammatory orchitis ( Fig 8 ) . In addition , we observed that the damaged testes of PRVABC59-infected mice had not recovered at 6 months post-inoculation ( S2 Table ) , which suggests that in humans , ZIKV infection might affect reproductive ability over an extended period . However , a recent report indicated that ZIKV infection produced no clear adverse effect on morphology or hormonal production in human testis explants [43] . Moreover , no overt testis damage was observed in a primate model of ZIKV infection , suggesting that the integrity of the host immune response might be associated with the sensitivity of the testis to ZIKV infection [44] . Comparison of the amino acid sequences of Asian/American-subtype ZIKV strains revealed differences in several amino acid residues between the subtypes ( Fig 9 ) . The amino acid at position 139 of prM is a serine ( 139S ) in the Southeast Asian subtype , but an asparagine ( 139N ) in the Pacific and American subtypes . This site has been shown to be involved in the ZIKV proliferative activity in neural progenitor cells that causes microcephaly in mice; however , a recent report indicated that a confirmed congenital ZVD case with microcephaly in Thailand was caused by infection with a Southeast Asian-subtype virus harboring 139S [18 , 45] . In contrast , our current results show that the Pacific and American subtypes of the virus produced clearly different pathogenic effects on the testis . However , there are distinct differences between the biological system of the fetal brain and male reproductive organs; thus , pathogenesis caused by ZIKV infection in the fetal brain does not necessarily equate with that in male reproductive organs . Therefore , factors such as variations in NS3 , NS5 , and other regions , which differ among the three strains , might be associated with ZIKV proliferation and the pathogenic effect on the testis and other male reproductive organs ( Fig 9 and Table 2 ) [17 , 46] . Point-by-point analysis performed using a ZIKV reverse-genetics system is required in order to identify the site ( s ) responsible for efficient proliferation and persistent infection of ZIKV in male reproductive organs . In this study , we used only one strain for each ZIKV subtype . Several variations might exist among the strains in each subtype and the strains used in the study might not necessarily reflect the properties of each subtype . Recent reports have indicated that Southeast Asian-subtype strains do not always show lower growth potential in cultured cells or lower virulence in mice compared with American-subtype strains , although the Southeast Asian-subtype strains that were used in these studies were repeatedly passaged in mouse brains [47 , 48] . In contrast , Smith et al . showed that low-passage Southeast Asian-subtype strain PHL/2012/CPC-0740 exhibits high pathogenicity in IFNAR1-KO mice [49] . Thus , further evaluation must be conducted using an increased number of ZIKV strains for each subtype to elucidate the differences in characteristics among the ZIKV subtypes . | Zika virus ( ZIKV ) is classified into two lineages , African and Asian/American . Phylogenetic analyses have revealed that Asian/American-lineage ZIKV strains can be divided into three distinct subtypes , the American , Pacific , and Southeast Asian subtypes , presenting several amino acid differences . In this study , we examined the in vitro and in vivo growth of three Asian/American lineage ZIKV strains belonging to the three subtypes . The American-subtype strain and the Southeast Asian-subtype strain exhibited the highest and lowest growth potential in vitro , respectively , and mice infected with these ZIKV strains also showed the highest and lowest viremia levels and infectious virus levels in the testis . Moreover , the rate and extent of testis damage were highest in mice infected with the American-subtype strain . Our results indicate that the American-subtype and Pacific-subtype strains exhibit increased ability to grow in vitro and in vivo and to induce testis damage in mice . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"sequencing",
"techniques",
"taxonomy",
"vero",
"cells",
"medicine",
"and",
"health",
"sciences",
"reproductive",
"system",
"pathology",
"and",
"laboratory",
"medicine",
"pathogens",
"biological",
"cultures",
"microbiology",
"viruses",
"amino",
"acid",
"sequence",
"analysis",
"phylogenetics",
"data",
"management",
"rna",
"viruses",
"phylogenetic",
"analysis",
"epididymis",
"molecular",
"biology",
"techniques",
"research",
"and",
"analysis",
"methods",
"sequence",
"analysis",
"computer",
"and",
"information",
"sciences",
"bioinformatics",
"medical",
"microbiology",
"microbial",
"pathogens",
"cell",
"lines",
"evolutionary",
"systematics",
"molecular",
"biology",
"testes",
"dna",
"sequence",
"analysis",
"anatomy",
"flaviviruses",
"viral",
"pathogens",
"database",
"and",
"informatics",
"methods",
"biology",
"and",
"life",
"sciences",
"evolutionary",
"biology",
"protein",
"sequencing",
"organisms",
"zika",
"virus",
"genital",
"anatomy"
] | 2019 | Increased growth ability and pathogenicity of American- and Pacific-subtype Zika virus (ZIKV) strains compared with a Southeast Asian-subtype ZIKV strain |
To understand the process of innate immune fungal recognition , we developed computational tools for the rigorous quantification and comparison of receptor recruitment and distribution at cell-cell contact sites . We used these tools to quantify pattern recognition receptor spatiotemporal distributions in contacts between primary human dendritic cells and the fungal pathogens C . albicans , C . parapsilosis and the environmental yeast S . cerevisiae , imaged using 3D multichannel laser scanning confocal microscopy . The detailed quantitative analysis of contact sites shows that , despite considerable biochemical similarity in the composition and structure of these species' cell walls , the receptor spatiotemporal distribution in host-microbe contact sites varies significantly between these yeasts . Our findings suggest a model where innate immune cells discriminate fungal microorganisms based on differential mobilization and coordination of receptor networks . Our analysis methods are also broadly applicable to a range of cell-cell interactions central to many biological problems .
C . albicans is a commensal of the human oropharyngeal cavity , gastrointestinal tract and female lower reproductive tract . It is also a significant opportunistic pathogen [1] . Infection by Candida species causes illnesses ranging from superficial mucosal infections that markedly diminish quality of life to bloodstream infections associated with high mortality . Systemic fungal infections by C . albicans have emerged as important causes of sickness and death in immunocompromised patients [2] . Some major risk factors associated with Candidemia involve neutropenia and prolonged hospitalization ( days ) involving in-dwelling medical devices which can become infected with Candida [3] . There is mortality rate associated with systemic Candida infection and an increased incidence of these types of infections in cancer patients [4]–[6] . For instance , Candida accounts for about one quarter of the fungal infections seen in leukemia patients [7] . During tissue colonization and invasion , C . albicans can undergo a transition from ellipsoidal yeast to filamentous hyphae , and this dimorphism is thought to be important for the infectious process . C . parapsilosis is one of the more commonly isolated non- albicans Candida species and is particularly problematic in neonates . It is clinically identified in 7–21% of systemic Candidiasis cases , where it is associated with 10–28% mortality [8]–[10] . C . parapsilosis colonizes human skin and nails , which is significant for its role in nosocomial infection [11] . C . parapsilosis can also be isolated from non-human animals , soil and physical surfaces [12] . S . cerevisiae is an environmental yeast most commonly associated with baking and fermentation processes . It is an exceedingly rare human pathogen , but can infect severely immune compromised patients [13] . The differing lifestyles of the three species compared may require different adhesive properties and regulation of cell wall structures so these fungi may adapt to and persist within their various niches . Nevertheless , they all contain grossly similar cell wall polysaccharide components and organization . Around 85% of the C . albicans cell wall is made up of diverse carbohydrates—primarily mannoproteins , -glucans , and chitin [14]–[16] . Chitin is deposited at sites deep within the cell wall and also exhibits some surface-accessibility at yeast bud scars [17]–[19] . However , the outermost layer of the Candida cell wall presents an external surface dominated by N-linked glycans which are comprised mostly of mannans [20] with punctate exposure of - and -glucans [17]–[19] . The cell wall contains a variety of mannosylated species including protein N- and O-linked -mannosides [16] , β-linked mannosides within N-linked mannan [21] and phospholipomannan [22] , [23] . Cell wall polysaccharides are essentially immobile on the time scale of host-pathogen interaction . Candida may modulate the degree of ligand exposure during infection [24] . Because the fungal cell wall is so complex , leukocytes must use multiple receptors in order to detect , interact with and initiate immune responses to fungal pathogens [20] , [25] , [26] . Innate immune cells , such as dendritic cells ( DCs ) , rely on pattern recognition receptors ( PRRs ) to identify fungal pathogens . These PRRs recognize pathogen-associated molecular patterns , which are characteristic molecular signatures of microbial biology [1] , [27] , [28] . Significant PRRs for fungal mannan recognition include the C-type lectins ( CTLs ) DC-SIGN , CD206 ( Mannose Receptor ) , Dectin-2 and Mincle ( N-linked mannan ) ; the Toll-like receptors TLR4 ( O-linked mannan ) and TLR2 ( phospholipomannan ) ; and Galectin-3 ( -linked mannosides ) [23] , [25] , [26] , [29] , [30] . -glucans are also immunogenic ligands of Dectin-1 ( a CTL ) and can be recognized by the integrin Mac-1 [31] . These receptors are expected to be relatively mobile in the plasma membrane . Recent research advances have clarified the identities of many receptors involved in fungal recognition , and increasingly ( i . e . , for DC-SIGN and Dectin-1 ) , signal transduction cascades have been elucidated [32] . For Candida albicans , there is evidence that receptors can tailor specific downstream signaling and cytokine responses depending on the morphological state of the pathogen . For example , investigators have reported that CLR-mediated recognition of both C . albicans yeasts and hyphae [33] , [34] and C . parapsilosis [35] results in divergent T helper cell polarization responses . Nevertheless , the specific contributions of individual receptors and their integration into the larger , multi-receptor system of fungal pattern recognition is not clear . Despite their ability to bind important pathogenic antigens , genetic ablation of CD206 or a murine homolog of DC-SIGN , SIGNR1 , has been shown to have little impact on host defense in murine models of Candidiasis and S . mansoni infection [36] , [37] . However , the existence of redundant systems for mannan sensing and species-specific differences in CTL function likely explain these findings . Furthermore , the interaction of Candida mannan with CD206 and DC-SIGN is well recognized as an important event in the generation of cytokine responses and phagocytosis by leukocytes [25] , [32] , [38]–[40] . While the functional consequences of CTL engagement are partially overlapping , evidence suggests that specific CTLs may be important for specific functions such as pathogen binding , phagocytosis and inflammatory cytokine generation [41] and co-engagement can modify CTL function [42] . Innate immune antigen presenting cells , such as dendritic cells , are some of the first responders to fungal infections and they also activate adaptive immune responses that are critical for clearing Candida infections [43] , [44] . The earliest event that occurs in response to a Candida infection is the formation of a contact between an innate immune cell and the pathogenic fungal cell , which then determines the course of downstream signaling to activate inflammatory responses . Understanding the biology of fungal recognition requires elucidation of 1 ) the transport of C-type Lectins and other pattern recognition receptors to the site of host-microbe interaction , 2 ) rearrangement and coalescence of these receptors to achieve lateral segregation or clustering , and 3 ) the initiation of signaling cascades at the host-microbe contact site . Despite the identification of various receptors involved in fungal recognition , many questions remain regarding the mechanisms of receptor assembly at host-fungal pathogen contact sites , the role of receptor aggregation at nano- and micrometer length scales [45] , and the spatiotemporal regulation of receptor cross-talk [31] , [46] . Key to answering these questions are tools that provide rigorous quantification of receptor redistribution and signaling at host pathogen contacts . The distribution of CTLs can be imaged at high resolution by three-dimensional multicolor confocal laser scanning microscopy ( 3D CLSM ) . A major difficulty in developing analysis tools is that the imaging data is collected using rectangular voxels while the yeast cell is nearly spherical and rigid , so the contact between the yeast and dendritic cell is part of an essentially spherical surface ( Fig . 1 ) . To overcome this difficulty , we developed geometric algorithms that construct spherical voxels that contain the yeast cell . The intensities in the rectangular voxels are transferred to the spherical voxels and then projected onto the surface of a sphere that approximates the surface of the yeast cell using weighted sums along the radial direction . The approximation is lenient , so a spectrum of geometries of the contact site are tolerable as long as they reside on a roughly spherical surface or within a spherical shell . We used these tools to quantitatively compare the differences in the contact site organization for the pathogens C . albicans , C . parapsilosis , and the environmental yeast S . cerevisiae . Some previous studies have used spherical coordinates to analyze biological data in ways that are related to , but significantly extended by , what we do here [47]–[50] . For instance , the tool we describe solves the above problems with particular attention to accurate transfer of intensity information to spherical voxels , use of equal area surface pixels for orientation-independence of contact site quantification , and a user-friendly interface that automatically computes a variety of spatial statistical measurements to assist in analysis of cell-cell contacts .
We cultured immature DCs with yeast cells for various times , then fixed the cells and fluorescently labeled the CTLs , DC-SIGN and CD206 , as well as the DC membrane lipids , as described in Materials and Methods . We used one environmental yeast ( Saccharomyces cerevisiae ) and two pathogenic yeasts ( Candida albicans and Candida parapsilosis ) to form the host-microbe contacts . We have chosen to focus our attention on these fungi because Saccharomyces and Candida cell wall composition and structure are thought to be mostly similar ( see Discussion ) , yet the innate immune system is often called upon to discriminate between harmless environmental fungi and pathogenic ones . Furthermore , we have focused on two receptors prominently involved in mannan recognition in order to elucidate how mannan sensing is orchestrated . Three color 3D fluorescence distributions at cell-pathogen contact sites were measured by 3D CLSM . Representative examples of the initial data are shown in Fig . 2 . We compared DC-SIGN and CD206 at fungal contacts formed in response to S . cerevisiae , C . albicans and C . parapsilosis with respect to spatiotemporal patterns of receptor entry at 0 , 1 and 4 hours of exposure to yeasts . These time points were chosen to focus on stable contact sites . Previous research has shown that the majority of zymosan particles bound to human DCs exhibit stable extracellular contacts over hours , and CTL signaling can occur from extracellular contacts with fungal ligands and from internal compartments over prolonged periods of time [51]–[53] . We observed differential CTL spatiotemporal distribution patterns in contact sites with the three fungal species . These contact sites contained zones that were colocalized ( on a diffraction limited scale ) or single positive ( schematically represented in Fig . 3A ) . S . cerevisiae and C . albicans provoked the greatest amount of DC-SIGN and CD206 recruitment respectively , within the first hour , and then both lost receptor intensity in the fourth hour . In contrast , C . parapsilosis continued to recruit significant amounts of both receptors from the start of the experiment into the fourth hour ( Fig . 3B , C ) . The slower recruitment of DC-SIGN by C . parapsilosis resulted in contact site accumulations that were times less than S . cerevisiae and times less than C . albicans at the first hour ( Fig . 3D ) . However , by the fourth hour , C . parapsilosis had recruited times more DC-SIGN than C . albicans and was still significantly less than S . cerevisiae ( Fig . 3E ) . Similarly , C . parapsilosis recruited CD206 slowly , times less than both the other yeasts ( Fig . 3F ) , but by the fourth hour recruited times more than S . cerevisiae and times more than C . albicans ( Fig . 3G ) . We observed large increases in DC-SIGN intensity recruited to the contact site in the first hour for S . cerevisiae , C . albicans and C . parapsilosis : 161-fold , 140-fold and 82-fold , respectively . Likewise , we observed contact site enrichments , albeit lower in magnitude , for CD206 intensity in the first hour for S . cerevisiae , C . albicans and C . parapsilosis : 63-fold , 73-fold and 34-fold , respectively . This data suggested that DC-SIGN and CD206 recruitment patterns varied in a manner that was quite sensitive to the species of yeast being recognized by the DC—both in terms of the amount and spatiotemporal distribution of receptor recruited . It was further notable that DC-SIGN , and CD206 to a somewhat lesser extent , was highly enriched in contact sites relative to resting cells and that both CTLs were well recruited to C . albicans contacts , as seen for the other yeasts as well . Receptor total intensity increase might derive from an increase in contact site area and/or increase of receptor density in contact sites . We proceeded to examine the contribution of these factors , starting with an assessment of contact site area . For all cases , we found that augmentation of CTL contact site area occurred most dramatically in the first hour , which is expected based on previously reported findings with macrophages interacting with C . albicans [54] . We found significant differences in the evolution of contact site area for DC-SIGN and CD206 amongst the three fungal species used to challenge DCs . S . cerevisiae was notable for the fact that it produced the contacts with largest area occupied by either receptor over the course of the experiment ( Fig . 4A , B ) . In contrast , both C . albicans and C . parapsilosis contacts were significantly smaller at one hour for both individual CTL contact site areas and total contact area ( Fig . 4A , B ) . S . cerevisiae contacts contained at least and times larger DC-SIGN and CD206 area than either of the other yeasts at one hour ( Fig . 4C , D ) , and times greater DC-SIGN and CD206 area relative to C . albicans at four hours ( Fig . 4E , F ) . S . cerevisiae contacts rapidly and effectively expanded , likely indicating a strong cytoskeletal response driving pseudopod extension for engulfment of the yeast . In contrast , C . albicans failed to produce contact site areas comparable to S . cerevisiae at either time point ( Fig . 4A , B ) . This may reflect a blunted cytoskeletal response to C . albicans and poorer engulfment , which is addressed further below . While S . cerevisiae and C . albicans contacts were quantitatively different but followed a similar pattern of CTL spatiotemporal distribution , C . parapsilosis contacts were qualitatively different from the other yeasts contacts in that they exhibited a slow , progressive area increase ( Fig . 4A , B ) . This progressive area increase for C . parapsilosis mirrored a similar trend seen for receptor recruitment ( Fig . 3B , C ) . To address the question of whether contact site area correlated with fungal particle size , we measured the major and minor radii of S . cerevisiae , C . albicans and C . parapsilosis yeasts ( each ) from DIC images ( data not shown ) . From these measurements we also calculated mid-sectional elliptical perimeters . Upon comparing these results by ANOVA and post-hoc test , we determined that C . albicans and S . cerevisiae yeast sizes were not significantly different for any of these quantities . C . parapsilosis did exhibit significantly larger major radii ( ) and elliptical perimeters ( ) compared to S . cerevisiae . S . cerevisiae generated the largest contact sites and C . albicans had the smallest contacts , yet these yeasts were similar in size . Therefore , we conclude that contact site size is not dictated by particle size but is more likely a reflection of the DCs response to the particle . Next we wanted to examine what population of the CTLs contributed to the increase in area . As illustrated in Fig . 3A , the contact can be divided into membrane regions with receptors that are colocalized at the limit of resolution ( , ) and single positive ( , ; or , ) regions . After analyzing the different populations of CTLs within the contact site , we found that the significant increase in total receptor area was primarily due to an increase in colocalized populations of CTLs in the contact site ( Fig . 4G ) . On the contrary , both populations of single-positive CTLs ( DC-SIGN and CD206 ) did not change significantly throughout the experiment and comprised a small fraction of the total contact site ( Fig . 4H , I ) . Taken together , our observations demonstrate that the spatial assembly of the contact site structure is regulated differentially in response to the fungal species presented . It is also clear that all examined contact sites prominently featured increased predominance of receptor-colocalized membrane areas . Notably , C . albicans recognition by DCs generated the smallest contact sites despite our finding that this yeast was not deficient in recruiting DC-SIGN or CD206 total intensity . The clustering of receptors at cell-cell contacts is a common theme in immunoreceptor signaling , and this mechanism drives the formation of membrane regions with increased receptor density . Receptor density is one factor that can regulate the efficiency of signal transduction and membrane trafficking of the receptor . Because receptor density in the contact is coordinately defined by the total amount of receptor recruited and the membrane area that it occupies , we created density graphs to display the difference between colocalized and single-positive DC-SIGN and CD206 distributions ( Fig . 5 ) . Fig . 5A provides a schematic example of contact site density over three time points ( T1-3 ) , and Fig . 5B provides the corresponding density graph analysis . At T1 , there is a small area with a small amount of intensity within that area that increases in intensity but not area in T2 ( thus , higher density in T2 vs . T1 ) . At T3 , this region exhibits increases in area and intensity . The dashed “isodensity” line depicts the set of all combinations of intensity and area with the same density as at T2 . Thus , because the slope is greater than that of the isodensity line ( i . e . , T3 lies in the green shaded area ) , the transition involves an increase in density at T3 relative to T2 . This would not be immediately apparent without reference to the isodensity line . In colocalized regions ( where both DC-SIGN and CD206 are found within the same voxel ) , we found that C . albicans accumulated the highest density for both DC-SIGN and CD206 within the first hour ( Fig . 5C , D ) . The same trend was also found in S . cerevisiae and C . parapsilosis , but with somewhat lower CTL densities achieved ( Fig . 5E , F , G , H ) . The development of a pronounced colocalized region with high receptor density could promote receptor cross-talk and strong adhesion . When we compared fungal species to one another , we found that C . albicans accumulated times more colocalized DC-SIGN density than S . cerevisiae and C . parapsilosis at the first hour ( Fig . 5C , E , G ) , but interestingly C . albicans accumulated times more colocalized CD206 than S . cerevisiae and C . parapsilosis ( Fig . 5D , F , H ) . Contact sites with S . cerevisiae and C . albicans both reduced their CTL colocalized density between the first hour and fourth hour ( Fig . 5C , D , E , F ) , whereas C . parapsilosis likewise gained density but did not exhibit an area or intensity loss at longer duration ( Fig . 5G , H ) . We note that all contacts increased their receptor density greatly in the first hour ( slopes well above the stated isodensity line ) , but C . albicans contacts were notable for being dense because they recruited DC-SIGN and CD206 well but remained small in area . Prior to our detailed analysis of the contact sites , we used the Manders coefficients to estimate the degree of colocalization . The coefficient M1 ( the proportion of DC-SIGN colocalized with CD206 ) indicated very high degrees of colocalization in 1 and 4 hour contacts for all three yeast species and both CTLs . As the Manders coefficients are influenced by both degree of overlap and intensity , they are not completely specific for variations in the amount of colocalization . Our contact site analysis provides more detailed results on colocalization in general . In this case , the Manders analysis and our contact site analysis of colocalization agreed with one another in finding predominant colocalization in contacts under all tested conditions . We hypothesized that the differential spatiotemporal patterns of receptor recruitment that we observed for S . cerevisiae , C . albicans , and C . parapsilosis would be correlated with the functional differences in binding and/or phagocytic efficiency during DC-yeast interaction . In particular , the smaller area contacts observed for C . albicans were suggestive of less actin reorganization and pseudopod extension . We quantified binding and phagocytic efficiency for DCs treated with yeasts for 1 and 4 hours , as described in the methods section . Interestingly , there was no significant difference in the median number of yeasts captured per DC between species at 1 or 4 hours ( Fig . 6A , B ) . We categorized DCs based on their interaction with yeasts as “neither” ( no bound or internalized yeast; excluded from analysis ) , “bound” ( only surface bound yeast ) , “internalized” ( only internalized yeast ) , and “B&I” ( some bound and some internalized yeasts ) . Despite this equivalent capture of yeasts , we found that DC populations exposed to C . albicans were skewed to distributions that reflected lower levels of internalization ( i . e . , decreased percent of the population in the “B&I” category ) relative to that seen for DCs exposed to S . cerevisiae or C . parapsilosis ( Fig . 6C , D ) . To understand this phenomenon in more detail , we examined cumulative probability distributions of phagocytic efficiency ( PE ) for DCs exposed to all three yeasts over 1 or 4 hours . We found that the proportion of DCs that failed to internalize any bound yeast ( ) was higher for C . albicans than the other species for both time points ( Fig . 6E , F ) . Furthermore , of those DCs that did internalize some yeasts ( ) , these DCs exhibited generally lower phagocytic efficiencies for C . albicans than other species . These trends represented a significant difference in PE distributions for C . albicans versus S . cerevisiae at 1 and 4 hours , and a significant difference between C . albicans and C . parapsilosis at 4 hours . The distribution of PE values was not significantly different between S . cerevisiae and C . parapsilosis at either time .
The analysis tool that we developed allows quantification of receptor behavior on an approximately spherical surface extended across multiple -axis confocal sectioning depths . This capability , coupled with the ability to resolve and quantify receptor structures on this host-pathogen contact site surface , allowed us to discern interspecies differences in CTL mobilization and organization during fungal recognition by dendritic cells . Despite the presence of abundant -mannoside ligands of DC-SIGN and CD206 in the cell walls of all fungi tested , we observed dissimilar spatiotemporal patterns of receptor recruitment amongst S . cerevisiae , C . albicans and C . parapsilosis . DCs recruited DC-SIGN and CD206 to contact sites with all three yeast species to achieve tens to over a hundred fold enrichment of receptors . However , receptor recruitment peaked earlier for C . albicans and S . cerevisiae , while C . parapsilosis contacts developed in a slower , progressive manner . Also interesting was the observation that S . cerevisiae contacts were quite large while C . albicans contacts were notable for being the smallest at both one and four hours . Because contact site area is likely to reflect the success of cytoskeletal remodeling in response to fungal recognition , we examined whether receptor recruitment patterns or contact site area characteristics correlated with the functional outcome of phagocytosis . We found that , despite similar ability to capture all yeasts , DCs exhibited significantly lower phagocytic efficiency when challenged with C . albicans in comparison with S . cerevisiae and C . parapsilosis . These data suggest that strong contact site recruitment of mannan-binding CTLs is important for capture of fungi by DCs , which is consistent with the fact that mannan is the dominant ligand on the cell wall surface . However , intensity of DC-SIGN or CD206 recruitment is not a strong predictor of phagocytic outcome . For instance , S . cerevisiae recruited the most DC-SIGN at one hour , while the intensity of DC-SIGN in C . parapsilosis contacts was much slower to develop to similar levels , but both yeasts were well-phagocytosed with similar efficiencies . Contact site area was a good predictor of phagocytic efficiency , and it is likely that both readouts reveal a relative paucity of cytoskeletal response to C . albicans yeast relative to S . cerevisiae or C . parapsilosis . This could reflect the existence of cell wall features possessed by C . albicans that minimize phagocytosis and aid in partial evasion of the innate immune response . These differences in spatiotemporal distribution patterns may result from subtle differences in the fine structure of mannan . C . albicans mannans have been shown to contain structural features such as β- ( 1 , 2 ) -linkages and branching α-linked oligomannoside side chains [55] , [56] which are not shared by S . cerevisiae or C . parapsilosis . Mannan structural differences can influence the antigenicity and surface chemistry of the cell wall [57] , [58] . In contrast to other cell-cell contact signaling systems with more laterally mobile ligand/receptor pairs ( i . e . , the immunological synapse ) , the ligands presented by the fungal cell wall are part of a dense and highly interconnected network . Although the cell wall does undergo remodeling , the lateral mobility of polysaccharide ligands in the contact site is quite low . Interestingly , recent work from Dufrêne and Lipke and colleagues has demonstrated that important mannoproteins of the Als adhesin family can be reorganized into distinct 100–500 nm amyloid domains in the cell wall of C . albicans upon application of force , and changes in Als protein exposure and organization are also seen under conditions such as hyphal germination and treatment with echinocandin drugs [59]–[62] . The consequent spatial reorganization of mannan ligands could be important for the nanoscale organization of DC-SIGN and CD206 in contact sites with DCs . Als adhesins are anchored to fibrillar glucan in the cell wall and above referenced results suggest that their mobility in the cell wall consists of gyration about their anchorage points , not long-range lateral mobility . However , some mannoproteins are known to be non-covalently associated with the cell wall and these could possess greater lateral mobility . In our analysis of fungal contact sites , we saw that receptors congregated in specific , micron-scale membrane structures despite presumed low levels of ligand lateral mobility . This study utilized fixed yeasts to provide more controlled experimental conditions and more straightforward data interpretation . This simplification precludes mannoprotein mobility during DC-yeast interaction , so future experiments in live cell interaction systems will be necessary to fully elucidate the role of fungal cell wall reorganization in these host-microbe interactions . The organization of receptors into micron-scale membrane substructures , wherein transmembrane protein populations may mix and achieve altered density , will likely influence the efficiency and maintenance of signal transduction . A previous report describing the “phagocytic synapse” showed that the lateral reorganization of the CTL Dectin-1 and the phosphatase CD45 influences Dectin-1 signaling [63] . The mechanisms that drive the formation of specific membrane structures in fungal contacts , such as ligand patterning on cell wall surfaces , observed for patches of -glucan exposure on C . albicans [24] , [64] , are an interesting topic for future research . CTLs have been described to exist in DC membranes as discrete nanodomains of approximately 80–100 nm diameter by several imaging methods such as transmission electron microscopy , near-field scanning optical microscopy and super resolution fluorescence imaging [45] , [65]–[68] . These domains have interesting biophysical properties , such as a lack of exchange of receptor with the surrounding membrane and nearly complete segregation of DC-SIGN and CD206 nanodomains in resting DC membranes [68]–[70] . Recently , we have observed that nanoscale organization of CTLs in fungal contacts is altered relative to non-contact membrane in favor of less individual nanodomain structure and more longer-range nanostructure , consistent with close packing of domains ( unpublished data , AKN ) . The significance of receptor colocalization and changes in receptor density in contact sites is that spatial proximity influences signal transduction by increasing amplitude and persistence of signaling as well as promoting crosstalk between receptors . Application of our analysis tool to higher resolution imaging modalities , such as Stimulated Emission Depletion microscopy and 3D direct Stochastic Optical Reconstruction Microscopy , may provide insights into critical early receptor rearrangement events in innate immune fungal recognition in future studies . Cell-cell contacts are a common theme in biology , being integral to such diverse processes as lymphocyte activation , tissue development and neural communication . Therefore , we anticipate that this tool will have broad utility in other fields where quantification of receptor and/or organelle mobility relative to a cell-cell contact is needed . Some examples of other potential biomedical applications include other phagocytic synapses ( i . e . , macrophage scavenging of apoptotic bodies ) , the immunological synapse between T cell and antigen presenting cell , receptors within the synapse between neurons , the association between plasma membrane and SNARE complexes on the ER for calcium signaling , between cytotoxic T cells or NK cells and virally infected target cells , and B cell or mast cell activation by particulate antigen . Much information can be derived from standard confocal optical imaging , as we have demonstrated . However , promising progress in techniques for 3D super resolution microscopy should provide access to structural detail on at least a log-order higher resolution , and such data could be analyzed by our method to assess changes in biologically significant structures such as receptor microclusters and STIM/Orai mediated signaling microdomains .
C . albicans ( ATCC , Manassas , VA , #MYA-2876 ) , C . parapsilosis ( ATCC , Manassas , VA , #22019 ) , and S . cerevisiae ( ATCC , Manassas , VA , #26108 ) were cultured in YPD broth in an orbital incubator at until exponential phase growth . Prior to application to dendritic cells , yeasts were fixed with 2 . 5% PFA at room temperature for 20 min followed by extensive PBS washing . We obtained human peripheral blood leukocytes from discarded leukocyte reduction filters provided by United Blood Services of Albuquerque . The filters were back-flushed with 300 mL HBSS , and the collected cells were spun over Ficoll-Paque Plus ( GE Healthcare , Sweden , #17-1440-02 ) . Monocytes were purified by adherence on tissue culture flasks . Immature dendritic cells were prepared by differentiation of monocytes in RPMI supplemented with 10% FBS , 1% Penicillin/Streptomycin , 10 mM Hepes , and 1 mM sodium pyruvate , 500 IU/mL human IL-4 ( Peprotech , Rocky Hill , NJ , #200-04 ) and 800 IU/mL human GM-CSF ( Sanofi , Bridgewater , NJ , Leukine/sargramostim/ ) at , 5% for 7 days . Immature DCs existing in 7 day cultures were exposed to the specified yeasts ( per sample ) for the specified times . These conditions were found to represent a relatively light challenge for DCs with yeast that is unlikely to overwhelm the ability of DCs to bind yeast , recruit receptors to contact sites or engulf particles . This use of human blood products was reviewed and approved by the University of New Mexico Health Sciences Center Human Research Review Committee . Fixed specimens were blocked and stained with primary and secondary antibodies . Primary antibodies were as follows: anti-human CD209 ( BD Pharingen , San Diego , CA , #551186 ) and anti-MRC1 ( Abnova , Taiwan , #H00004360-M02 ) applied at a concentration of for 30 minutes at . These conditions provided an excess of primary and secondary antibodies and achieved saturation binding of receptors . Identical staining conditions were used in the preparation of all samples for contact site analysis . The following secondary antibodies were used: Alexa Fluor 488 goat anti-mouse ( Invitrogen , Grand Island , NY , #A21141 ) and Alexa Fluor 647 goat anti-mouse ( Invitrogen , Grand Island , NY , #A21240 ) applied at a concentration of for 30 minutes at . Cell membrane was visualized by Cell Mask Orange ( CMO ) ( Invitrogen , Grand Island , NY , #C10045 ) at a concentration of for 5 minutes at . This staining condition allows only DC membranes to stain . The CMO staining duration is insufficient to allow dye penetration of the cell wall for yeast plasma membrane staining . Fully phagocytized yeasts were not accessible to receptor staining and are thus not represented in contact site receptor analysis . Contact sites randomly chosen for analysis of receptor spatiotemporal distributions exhibited a range of expected engulfment morphologies . Fluorescent proteins and lipids were imaged with a FV1000 laser scanning confocal microscope ( Olympus , Center Valley , PA ) equipped with a 60× , 1 . 42 NA , Plan-Apochromat oil immersion objective . AlexaFluor488 ( reporting the distribution of CD209 ) was excited with a 15 mW , 473 nm diode laser operated at 1% power; AlexaFluor647 ( reporting the distribution of CD206 ) was excited with a 20 mW , 635 nm diode laser operated at 1% power , and CMO ( reporting the dendritic cell membrane 3D profile ) was excited with a 15 mW , 559 nm diode laser operated at 1% power . These lines were reflected to the specimen by a 405/473/559/635 multi-edge main dichroic element , and emission was routed through the main dichroic mirror and confocal pinhole ( 115 nm diameter ) to secondary longpass dichroics ( or a mirror ) followed by bandpass emission filters in front of 3 independent PMT detectors . Specifically , the emission light passed by the main dichroic was directed to PMT1 ( AF488/DC-SIGN channel ) via reflection from a longpass 560 nm cutoff dichroic mirror and passage through a BA490-540 nm bandpass filter . Emission passing this dichroic was directed to PMT2 ( CMO channel ) via reflection from a longpass 640 nm cutoff dichroic mirror and passage through a BA575-620 nm bandpass filter . Finally , emission light passed through this dichroic was directed to PMT3 ( AF647/CD206 channel ) via reflection from a mirror and passage through a BA655-755 nm bandpass filter . Z-stacks were recorded with 250 nm spacing . Other parameters were pixel dimensions ( square pixels in the dimension ) , pixel dwell time ( 2 µs/pixel ) , detector sensitivity ( PMT1 640 volts; PMT2 455 volts; PMT3 610 volts; gain = 1 and offset = 0 for all PMTs ) . All imaging parameters as described above were kept constant during acquisition of all images for contact site analysis . Photo bleaching was not found after examining the -axis profile of 0 hr time points , . Each experimental result presented in this work represents pooled data from independent replicates with DCs from three separate donors . Within any given replicate , DCs were chosen at random for imaging and analysis . We imaged a minimum of 50 contact sites per species per time point for contact analysis . Statistical significance was determined by ANOVA , Tukey post-hoc test . Fixed yeast were stained with two different markers prior to being added to the live DC culture . The first marker was Calcofluor White ( Sigma-Aldrich , St . Louis , MO , #F3543 ) at a concentration of for 20 minutes at . The second label was Biotin-NHS ( Sigma-Aldrich , St . Louis , MO , #H1759 ) , which was conjugated to cell wall proteins of yeasts ( fixed with paraformaldehyde , as above ) at a concentration of for one hour at in PBS at 8 . 5 pH . After staining , these yeast particles were added to live DC culture for either 45 minutes or three hours and 45 minutes . At either time point , streptavidin-Alexa Fluor 647 ( AF647 ) in RPMI warmed to was added to the live DC culture for 15 minutes . At this point the DCs were fixed with 4% PFA in PBS for 10 minutes at followed by extensive PBS washing . Fixed yeast particles were imaged with a FV1000 laser scanning confocal microscope ( Olympus , Center Valley , PA ) equipped with a , 1 . 42 NA , Plan-Apochromat oil immersion objective . Calcofluor White ( marker for all yeast ) was excited with a 50 mW , 405 nm diode laser operated at 1% power and streptavidin conjugated Alexa Fluor 647 ( marker for only external yeast ) was excited with a 20 mW , 635 nm diode laser operated at 1% power . These lines were reflected to the specimen by a 405/473/559/635 multi-edge main dichroic element , and emission was routed through the main dichroic mirror and confocal pinhole ( diameter ) to secondary longpass dichroics ( or a mirror ) followed by bandpass emission filters in front of 2 independent PMT detectors . Specifically , the emission light passed by the main dichroic was directed to PMT1 ( fluoresce brightener channel ) via reflection from the mirror and passage through a BA430-455 nm bandpass filter . Emission passing this dichroic was directed to PMT3 ( streptavidin conjugated to AF647 channel ) via reflection from a mirror and passage through a BA655-755 band pass filter . Z-stacks were recorded with spacing . Other parameters were voxel dimensions ( voxels in dimensions ) , pixel dwell time ( ) , detector sensitivity ( PMT1 650 volts , PMT3 570 volts ) ; gain = 1 and offset = 0 for all PMTs ) . All imaging parameters as described above were kept constant during acquisition of all images for phagocytosis assay . Bound and internalized yeast were enumerated manually on a per DC basis in all 3D confocal datasets . Bound yeasts were identified based on their location on DCs ( DIC ) and positive signal for both Calcofluor White and AF647 emission . Internalized yeasts were identified by apparent localization inside a DC ( DIC ) emission in the Calcofluor White channel only . We calculated the median and interquartile range for both categories over all DCs imaged . Phagocytosis Efficiency ( PE ) for each DC was calculated as the number of yeasts that were identified as internalized divided by the total number of yeasts associated with the same DC ( that is , surface bound plus internalized yeasts ) . Statistical significance was determined by the Mann-Whitney test . To facilitate the quantitative analysis of the contact sites , we developed a graphical user interface for the analysis programs . This interface allows the user to load the image files , specify parameters and select regions of interest , for example , see the first row in Fig . 2 . The images in this row show a few dendritic cells interacting with yeast cells . The user selects a yeast cell for analysis by clicking on it , which spawns a new window with a close-up view of the selected region . In the image of a single yeast cell , the user selects the center of the yeast and an inner and outer radius such that the yeast cell wall surface lies between the spheres determined by the two radii . The contact site is assumed to reside within this spherical shell and surfaces that depart significantly from sphericity ( e . g . , nearly planar regions ) can still be analyzed as long as the contact site falls within the spherical volume described by the two radii . The underlying analysis programs then transform the data to spherical coordinates and project the intensity values onto the outer spherical surface which approximates the yeast cell surface . The area of membrane/cell wall contact between the dendritic cell and yeast cell is identified by thresholding , and receptor fluorescence intensities and analysis results are written to a spreadsheet for further analysis . We have validated our method against artificial objects where the recovered receptor intensities , locations and colocalizations can be compared with our knowledge of the ground truth for these parameters . In more detail , our data sets have four channels: The red and green channels are intensities from two different fluorophores . For each channel , the data are the intensities of the light emitted in each voxel of a three dimensional image ( Z-stack ) . The transmitted light channel images of a few dendritic cells and several yeast cells are shown in Fig . 2 . These images are used to select a single yeast in contact with a dendritic cell that is to be analyzed . Using the interface , the user selects the center of the yeast and then draws two radii , and , that determine two spheres such that the voxels between the two spheres contain all of the light emitted from the contact site . To analyze these data , the program establishes spherical coordinates ( see Fig . 7 ) with origin at the center of the yeast cell . These coordinates are used to divide the space up into spherical voxels , and additionally divide the surface of the larger of the two spheres ( with radius ) into pixels as shown in Fig . 8A . We analyze the data for each color by first transferring the intensities from the rectangular to the spherical voxels . This is done by dividing all of the relevant rectangular voxels into many much smaller rectangular subvoxels , and then apportioning the intensities among the subvoxels . For each subvoxel , the spherical voxel that contains the center of the subvoxel is determined , permitting the intensities from that subvoxel to be transferred to the appropriate spherical voxel . This transfer is computationally expensive , so several techniques were developed to make this process more efficient . See below for further details . The localization/colocalization analysis starts with a spherical projection of the intensities viewed as periodic data in a rectangle; see Fig . 8B . The analysis program uses thresholding to identify the contact site and performs several standard image processing techniques to prepare the images for further processing . The analysis program selects a region containing the contact site and then determines the spherical pixels occupied by receptors , presents the results to the user , and also writes a spreadsheet file that can be used for additional analyses . | Specialized cell-cell contacts are a common theme in cell biology . These structures increase sensitivity and specificity of cellular activation and information flow in contexts ranging from activation of immune responses to transmission of nerve action potentials . Candida species fungal pathogens are responsible for significant morbidity associated with mucocutaneous infections as well as mortality ( ) caused by bloodstream infections . The initial contact between innate immune cells and Candida results in a cell-cell contact between host and microbe . Leukocytes mobilize a network of receptors to these contact sites , and these receptors collaborate to recognize molecular patterns characteristic of microbial surfaces . Receptor recruitment , activation , and cross-talk are critical determinants of the evolution of signaling that directs the activation of downstream immune responses . However , host-pathogen contacts with fungi are complex and variable , and accurate quantification of receptor distribution in space and time is difficult with existing image analysis tools . Therefore , we have developed computational algorithms and a user interface that allows the scientist to both visualize and quantify receptor distribution in and recruitment to cell-cell contacts . We have used this software to show significant differences in contact site receptor accumulation and organization for three different host-fungal contact sites with environmental and pathogenic fungi . We also explored the correlation of contact site characteristics with the important functional outcome of phagocytosis . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"biomacromolecule-ligand",
"interactions",
"immunopathology",
"biochemistry",
"protein",
"chemistry",
"immune",
"cells",
"mathematics",
"cell",
"biology",
"animal",
"cells",
"clinical",
"immunology",
"antigen",
"processing",
"and",
"recognition",
"antigen-presenting",
"cells",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"immunology",
"physical",
"sciences",
"biophysics",
"immune",
"response"
] | 2014 | A New Tool to Quantify Receptor Recruitment to Cell Contact Sites during Host-Pathogen Interaction |
Quinolones are potent broad-spectrum bactericidal agents increasingly employed also in resource-limited countries . Resistance to quinolones is an increasing problem , known to be strongly associated with quinolone exposure . We report on the emergence of quinolone resistance in a very remote community in the Amazon forest , where quinolones have never been used and quinolone resistance was absent in 2002 . The community exhibited a considerable level of geographical isolation , limited contact with the exterior and minimal antibiotic use ( not including quinolones ) . In December 2009 , fecal carriage of antibiotic resistant Escherichia coli was investigated in 120 of the 140 inhabitants , and in 48 animals reared in the community . All fluoroquinolone-resistant isolates were genotyped and characterized for the mechanisms of plasmid- and chromosomal-mediated quinolone resistance . Despite the characteristics of the community remained substantially unchanged during the period 2002–2009 , carriage of quinolone-resistant E . coli was found to be common in 2009 both in humans ( 45% nalidixic acid , 14% ciprofloxacin ) and animals ( 54% nalidixic acid , 23% ciprofloxacin ) . Ciprofloxacin-resistant isolates of human and animal origin showed multidrug resistance phenotypes , a high level of genetic heterogeneity , and a combination of GyrA ( Ser83Leu and Asp87Asn ) and ParC ( Ser80Ile ) substitutions commonly observed in fluoroquinolone-resistant clinical isolates of E . coli . Remoteness and absence of antibiotic selective pressure did not protect the community from the remarkable emergence of quinolone resistance in E . coli . Introduction of the resistant strains from antibiotic-exposed settings is the most likely source , while persistence and dissemination in the absence of quinolone exposure is likely mostly related with poor sanitation . Interventions aimed at reducing the spreading of resistant isolates ( by improving sanitation and water/food safety ) are urgently needed to preserve the efficacy of quinolones in resource-limited countries , as control strategies based only on antibiotic restriction policies are unlikely to succeed in those settings .
Quinolones are broad-spectrum antimicrobial agents with rapid bactericidal activity , overall low toxicity , and the possibility of being administered either orally or parenterally . Thanks to these features , quinolones are drugs of choice for the treatment of several community- and hospital-acquired infections ( such as respiratory tract infections , skin and soft-tissue infections , urinary tract infections , gastro-intestinal infections , gonorrhea , tuberculosis , etc . ) , being among the most prescribed antibiotics [1] . Moreover , despite pediatric use has been restricted due to concerns with bone cartilage toxicity , quinolones are increasingly prescribed also for the treatment of life-threatening infections in pediatric patients [2] , [3] . Following the broad dissemination of pathogens with acquired resistance to the older and less expensive antibiotics ( e . g . ampicillin , tetracycline and trimethoprim-sulfamethoxazole ) , and the recent release of low-cost generic ciprofloxacin , the consumption of quinolones has significantly increased also in resource-limited countries [4] , [5] . In those settings , where the newest and patent-protected antimicrobial compounds are not easily available , quinolones have become key drugs for the treatment of common bacterial infections , including those with a major impact in morbidity and mortality , such as dysentery and typhoid fever [6] , [7] . As observed with all other antimicrobial agents , also the quinolones are affected by bacterial resistance . Acquired quinolone resistance has been reported among all major bacterial pathogens , and has attained very high level rates in several important Gram-positive and Gram-negative pathogens ( including Staphylococcus aureus , Neisseria gonorrhoeae , Escherichia coli , Klebsiella pneumoniae , Salmonella enterica , Shigella spp . , Pseudomonas aeruginosa , Acinetobacter spp . , Helicobacter pylori ) in some settings [1] , [4]–[6] , [8]–[10] . Quinolone resistance generally arises in a stepwise manner , following chromosomal mutations that alter the topoisomerase targets or upregulate bacterial efflux systems . In Enterobacteriaceae , several plasmid-mediated resistance mechanisms to quinolones ( PMQR ) have also been detected , including the Qnr proteins ( that protect the topoisomerase targets ) , the AAC-cr enzyme ( that inactivates some quinolones by acetylation ) , and the QepA and OqxAB efflux systems ( which are able to extrude some quinolones ) [11] . Although these PMQR mechanisms are able to confer only low level resistance to quinolones , their presence is thought to facilitate the emergence of chromosomal mutations leading to resistance levels of clinical significance [11] . A clear relationship has been demonstrated between the emergence and dissemination of quinolone resistance among bacterial pathogens and fluoroquinolone use , both in hospital and community settings [5] , [12]–[14] , while a recent study has also reported a rapid decrease of quinolone resistance rates in clinical isolates of E . coli after a countrywide intervention of quinolone restriction [15] . The relationship between quinolone use and resistance has also been indirectly supported by the absence or very low rates of acquired resistance to these drugs in the few studies which investigated antibiotic susceptibility in enterobacteria isolated from humans or wild animals living in remote areas of the planet away from anthropogenic drug exposure [16] , [17] . Here we report on the experience with a very remote community of the Peruvian Amazon forest , where quinolone resistance in commensal E . coli was found to be completely absent in 2002 [18] , but present at remarkable rates in 2009 , notwithstanding that during this period the community had retained a condition of high level of geographical isolation , limited exchanges with the exterior , minimal antibiotic exposure and absence of quinolone availability .
Full ethical clearance was obtained from the qualified local authorities who had revised and approved the study design and consent form ( Comité Institucional de Ética de la Universidad Peruana Cayetano Heredia , Lima , Peru ) . Before the fieldwork started , representatives of the local healthcare authorities and the research team met the community leader and adults to explain the purpose and procedures of the survey . All the inhabitants of the community were considered eligible for the study . Prior to their enrollment , written informed consent was obtained from all adult participants and from the parents or legal guardians of minors . Any literate participant signed the consent form . In case of an illiterate participant , the consent form was read and signed by a witness ( who was present throughout the consent procedure and interview ) and marked by the participant/parent . Consent procedure and interviews were always conducted by trained local healthcare workers with the help of a local translator . Angaiza is a community of Chayahuita ethnic group located in the Alto Amazonas province of Peru . It was selected by the local healthcare authorities as being one of the most isolated community of the Peruvian Amazonas . In fact , from the nearest urban area ( Yurimaguas , about 32 , 000 inhabitants ) , Angaiza can be reached by a 13-hour trip , including a 2-hour drive on an unpaved road followed by a 4-hour motor boat ride and a final 7-hour walk in the jungle . The population lives in typical Amazon huts including a single room , without sanitation and electricity , and locally collected rainwater represents the only water source . The principal activities are agriculture , hunting and animal breeding ( poultry , pigs and cows ) . Healthcare available consists of the visits of a professional healthcare worker approximately every 4 months , and primary care for the most common illnesses provided by a volunteer from the community . A previous study on fecal carriage of antibiotic resistant enterobacteria among the inhabitants of Angaiza was performed in 2002 [18] . At the time of the 2009 survey , the Chayahuita community comprised 140 individuals , living in 21 households . One hundred twenty members of the community ( 86% ) consented to participate in the study ( female-to-male ratio 61∶59 , age range 0–71 years , mean age 17 years , median age 12 years ) ( Table 1 ) . Study participants were representative of all the 21 households of the community ( mean and median study participants per household was 6 and 6 , respectively ) . The study was conducted during a two-day visit to the community ( December 12–13 , 2009 ) . Specially prepared forms were used to collect data from the community leader ( about the general characteristics and organization of the community ) and from each individual/legal guardian of children included in the study ( about travels outside the community and previous antibiotic use ) . For the microbiological investigation , a stool sample was collected from each individual who consented to participate in the study , and a fecal swab was obtained from each sample . Moreover , fecal swabs were obtained from 48 animals reared in the community , including poultry ( n = 19 ) , pigs ( n = 13 ) , dogs ( n = 8 ) , cattle ( n = 6 ) and cats ( n = 2 ) . Fecal swabs were stored in Amies transport medium ( Oxoid , Milan , Italy ) and transferred within 48 hours to the laboratory of Santa Gema Hospital of Yurimaguas . Fecal carriage of antibiotic resistant E . coli was investigated by a direct plating method , as described previously [18]–[20] . Briefly , each fecal swab was spread onto a MacConkey Agar No . 3 plate ( MCA ) ( Oxoid , Milan , Italy ) to yield uniform growth , and antibiotic disks were directly placed onto the seeded plate . After incubation at 37°C for 12–14 hours , plates were inspected for coliform growth , and inhibition zone diameters were measured and interpreted according to the previously described breakpoints [19] , [20] . Criteria for differentiating between dominant and subdominant resistant population were the same as in the previous survey [19] . Briefly , a growth inhibition zone absent or smaller than the breakpoint diameter was suggestive of the presence of a resistant dominant population , while isolated colonies growing inside a growth inhibition zone of any size were considered suggestive of the presence of a resistant subdominant population . Antibiotics tested included ampicillin , ceftriaxone , tetracycline , trimethoprim-sulfamethoxazole , chloramphenicol , streptomycin , kanamycin , gentamicin , amikacin , nalidixic acid , and ciprofloxacin ( Oxoid ) . All fecal samples positive for the presence of a coliform population resistant to ciprofloxacin ( 17 from humans and 11 from animals ) were streaked onto MCA plates supplemented with 5 µg/ml ciprofloxacin ( MCA-CIP ) . One bacterial isolate exhibiting the morphology typical of E . coli was collected from each plate and identified by the API20E system ( bioMérieux , Marcy l'Étoile , France ) . Susceptibility testing was performed by the disk diffusion method according to Clinical and Laboratory Standards Institute ( CLSI ) [21] , [22] . E . coli ATCC 25922 was used for quality control purposes . Detection of PMQR genes ( qnrA , qnrB , qnrC , qnrD , qnrS , aac ( 6′ ) -Ib-cr , qepA ) was performed by PCR and sequencing , as described previously [23] . Sequence analysis of gyrA and parC was carried out as described previously [24] . Nucleotide sequences were determined on both strands of PCR amplification products at the Macrogen sequencing facility ( Macrogen Inc . , Seoul , Korea ) . Genotyping of ciprofloxacin resistant isolates was performed by determination of the main phylogenetic groups ( A , B1 , B2 , D ) using the Clermont method [23] , Random Amplification of Polymorphic DNA ( RAPD ) using the 1290 decamer [23] , and Multi Locus Sequence Typing ( MLST ) using protocols and conditions described on the E . coli MLST website [http://mlst . ucc . ie/mlst/dbs/Ecoli/documents/primersColi_html] . Data entry and analysis were performed with the Epi Info software package version 2008 ( Centers for Disease Control and Prevention , Atlanta , GA ) . Statistical differences were determined by the Chi-Squared test . Confidence intervals were calculated by Stata Software release 8 . 0 ( StataCorp . 2003 ) .
Data obtained by the community leader and participants interviews showed that the characteristics of Angaiza were overall comparable to those observed in a similar survey carried out in 2002 [18] ( Table 1 ) . In particular , no major changes of the population structure , habits and healthcare organization had occurred since 2002 . The most important difference consisted in the introduction of a panel of antibacterial drugs to be stored in the community ( absent in 2002 ) , including ampicillin , dicloxacillin , erythromycin and trimethoprim-sulfamethoxazole . Moreover , antimalarial drugs were no longer stored in the community , differently from 2002 when they had been introduced following a previous malaria epidemic ( 40% of individuals included in the 2002 study had received chloroquine in the two weeks preceding the survey [unpublished] ) . During the 12 months preceding the survey , 33 individuals ( 27 . 5% ) from 15 households had travelled to Yurimaguas ( the nearest urban area ) , revealing a similar mobility rate compared to that observed in 2002 ( Table 1 ) . As far as antibiotic use is concerned , in the two weeks preceding the survey antibiotics were administered to five children ( age range 0–5 years ) from three households , for the treatment of diarrheal diseases . In particular , three children had received ampicillin and two trimethoprim-sulfamethoxazole ( both drugs stored in the community ) ( Table 1 ) . Moreover , six individuals reported use of antibiotics in the 12 months preceding the survey ( excluding the last two weeks ) , although the type of antibiotic could not be identified . Despite the availability of some antibacterial compounds in the community , antibiotic use was found to be overall comparable to that observed in 2002 ( P = 0 . 76 and P = 0 . 19 for use in the 2 weeks or 12 months preceding the survey , respectively ) . The usage of antibiotics for veterinary , husbandry and agricultural practices , and the use of animal feed remained totally absent , as they were in 2002 . Of the 120 individuals included in the 2009 survey , 119 ( 99% ) were found to carry antibiotic resistant E . coli as part of their intestinal microbiota ( 92% in the 2002 survey , P = 0 . 008 ) . Compared to the previous survey [18] , the most relevant findings were the overall increase of resistance rates ( statistically significant for ampicillin , trimethoprim-sulfamethoxazole , streptomycin and kanamycin ) , and the emergence of resistance to quinolones at remarkable rates ( 45% to nalidixic acid , 14% to ciprofloxacin ) , which was completely absent in 2002 ( Table 2 ) . Of note , quinolone resistant E . coli represented the dominant enterobacterial population in a considerable proportion of individuals ( Table 2 ) . Carriers of quinolone resistant E . coli were found in 20 of the 21 households of the community . No significant association was found between carriage of quinolone resistant isolates and age , gender , travels to Yurimaguas ( or living in a household with at least one member reporting previous travels to Yurimaguas ) or antibiotic consumption ( or living in a household with at least one member reporting previous antibiotic use ) ( data not shown ) . Investigation of fecal carriage of antibiotic resistant E . coli in 48 animals reared in the community ( including poultry , pigs , cattle , dogs and cats ) showed resistance rates overall similar to that observed in humans ( Table 3 ) . Of note , resistance to quinolones was widespread in all studied animal species . All 28 ciprofloxacin-resistant E . coli isolates ( 17 and 11 of human and animal origin , respectively ) were investigated for susceptibility phenotype , genetic background , and mechanisms of plasmid- and chromosomal-mediated quinolone resistance . Ciprofloxacin-resistant isolates were always resistant to nalidixic acid and usually showed a multidrug resistance phenotype ( defined as resistance to >1 antibiotic class ) , which mostly included trimethoprim-sulfamethoxazole ( 89% ) , tetracycline ( 86% ) , ampicillin ( 79% ) , streptomycin ( 68% ) , and chloramphenicol ( 57% ) ( Table 4 ) . An overall genetic heterogeneity was observed among isolates of either human or animal origin . In fact , they were found to belong to various phylogenetic groups ( 61% group A , 28% group B1 , and 11% group D ) , and to 12 different RAPD types ( Table 4 ) . The three most prevalent RAPD types ( type A , including 6 isolates from 3 families; type B , including 6 isolates from 5 families; type G , including 6 isolates from 4 families ) were detected both in humans and animals and were assigned to ST617 , ST10 , and ST224 , respectively . Sequencing the QRDR regions showed the presence of a double substitution in GyrA ( Ser83Leu and Asp87Asn ) and a single substitution in ParC ( Ser80Ile ) in 27 ciprofloxacin resistant isolates , and of a double substitution both in GyrA ( Ser83Leu and Asp87Tyr ) and ParC ( Ser80Ile and Ala108Val ) in the remaining one ( Table 4 ) . None of the PMQR genes investigated was detected .
In a previous survey , conducted in 2002 , acquired quinolone resistance was found to be absent in E . coli from the inhabitants of Angaiza , a very remote community of Chayahuita ethnic group located in the Peruvian Amazonas and characterized by a considerable level of geographical isolation , limited contacts with the exterior , and minimal antibiotic use , which did not include quinolones [18] . In this study we showed that , despite the characteristics of the community remained substantially unchanged over a 7-year period , in 2009 quinolone resistant isolates were common in Angaiza , with carriage of nalidixic acid and ciprofloxacin resistant E . coli observed in 45% and 14% of the studied individuals , respectively . Quinolone resistant E . coli were also found to be common , at rates similar to those observed in humans , among animals of different species reared in the community . The genetic heterogeneity observed among ciprofloxacin resistant isolates of human and animal origin excluded that the emergence of quinolone resistance in Angaiza was the consequence of the occasional introduction into the community of a highly successful quinolone resistant clone , capable of spreading and persistence even in the absence of selective pressure . These findings rather supported the hypothesis of a consistent influx of resistant isolates into the community , despite remoteness and limited exchanges with the exterior , with their persistence and dissemination in the absence of quinolone exposure being favored by the conditions of poor sanitation confirmed by the finding of RAPD types shared by humans and animals of different households . The multidrug resistance phenotype expressed by most ciprofloxacin resistant isolates would be consistent with a provenance from urban areas where they are selected by antibiotic exposure . In fact , the substitutions in the QRDR regions of GyrA ( Ser83Leu and Asp87Asn ) and ParC ( Ser80Ile ) detected in the ciprofloxacin resistant isolates from Angaiza are known to be among the most common cause of acquired high level fluoroquinolone resistance in clinical isolates of E . coli worldwide [25] . An influx of resistant isolates from urban areas to the remote community was also hypothesized to explain the high resistance rates to the oldest antibiotics ( i . e . tetracycline , ampicillin , trimethoprim-sulfamethoxazole , streptomycin and chloramphenicol ) observed in Angaiza in the 2002 survey [18] . Evidences supporting this scenario were represented by the similarities of resistance patterns and resistance genes observed between Yurimaguas ( the nearest urban area ) and Angaiza . On that occasion , the lack of quinolone resistance in the remote community was thought to reflect the fact that quinolone resistance rates in Yurimaguas were lower than those to the oldest antibiotics . In this perspective , the emergence of quinolone resistance in Angaiza would be consistent with the dramatic increase of carriage of quinolone resistant E . coli observed among healthy children in Yurimaguas in the period 2002–2005 ( 27% vs . 54% and 16% vs . 31% for nalidixic acid and ciprofloxacin , respectively ) [20] , [26] . In a recent study conducted in very remote villages of rural Guyana not exposed to quinolones , the finding of ciprofloxacin resistant E . coli was putatively ascribed to the heavy exposure to chloroquine , an antimalarial drug that can select for topoisomerase mutations conferring resistance to quinolones [27] . This does not seem to be the case for the emergence of quinolone resistance in Angaiza , since this community is located in a malaria endemic area where chloroquine has been used for a long time ( e . g . 40% of individuals included in the 2002 study had received chloroquine in the two weeks preceding the survey ) , whilst quinolone resistance emerged only in recent years . Moreover , data from the 2009 interviews excluded a recent malaria outbreak in Angaiza and consumption of chloroquine in the two weeks preceding the survey , ruling out the hypothesis that the differences observed between 2002 and 2009 could have been related to an increased exposure to this antimalarial drug . table-4-captionFluoroquinolones have become increasingly important in the therapeutic armamentarium of resource-limited countries , following the availability of generics ( which drastically reduced drug costs ) and the dramatic increase of resistance to the oldest and cheapest antibiotic classes [4] , [5] . These broad-spectrum , stable and orally administrable antibiotics have entered among the first- and second-line choices for the treatment of several common bacterial infections , including enteric , respiratory and urinary tract infections , sexually transmitted diseases , as well as serious systemic infections ( e . g . typhoid fever , urinary sepsis , bacteremia in severe malnutrition ) [1] , [2] , [5]–[7] . Due to the frequent unavailability of the newer and patent protected antibiotics , the dissemination of fluoroquinolone resistance in resource-limited countries has worrisome clinical implications . Acquired quinolone resistance in enterobacteria has been clearly associated with the use of fluoroquinolones [5] , [12]–[14] , being absent or exceedingly rare in remote areas of the planet away from anthropogenic drug exposure [16] , [17] . Recently , a countrywide intervention of quinolone restriction in Israel resulted in a rapid decrease of resistance rates in clinical isolates of E . coli [15] , suggesting that maintenance of quinolone resistance in enterobacteria could be strongly dependent on drug exposure . Our study provided new insights into this phenomenon , as we demonstrated that quinolone resistant E . coli ( likely selected in urban areas under quinolone selective pressure ) were able to widely disseminate and persist even in very remote settings not exposed to antibiotics . These findings proved that maintenance of quinolone resistance in E . coli is not always depended on drug exposure , as also suggested by recent studies on fitness cost of quinolone resistance [28] , [29] , and emphasized the key role of the intestinal microbiota in the dissemination of such a clinically relevant antibiotic resistance . Overall , the results from the present study underline the urgent need for interventions aimed at improving sanitation and water/food safety to address the phenomenon of antibiotic resistance in resource-limited countries . Indeed , unless dissemination of resistant isolates is contained , control strategies based only on antibiotic restriction policies are unlikely to succeed in those settings , especially for bacteria able to colonize the human gut . | Quinolones are broad-spectrum antibiotics which bind to type II topoisomerases ( DNA gyrase and topoisomerase IV ) and inhibit DNA re-ligation after enzyme cut , exerting a rapid bactericidal activity . They are widely used for the treatment of several community- and hospital-acquired infections and have become increasingly important also in resource-limited countries , following the availability of generics ( which have drastically reduced drug costs ) and the remarkable increase of resistance to the oldest and cheapest antibiotic classes . Resistance to quinolones is an increasing worldwide problem that challenges the efficacy of these drugs against several bacterial pathogens and is known to be strongly associated with quinolone exposure . Restriction of quinolone consumption has been advocated as an important tool for the containment of quinolone resistance and has recently been proved to succeed in reducing resistance rates in clinical isolates of Escherichia coli in a community setting from an industrialized country . This study describes the dissemination of quinolone resistant E . coli in a very remote community in the Amazon forest , with a high level isolation and minimal antibiotic use , not including quinolones . These findings demonstrate that intervention strategies based only on quinolone restriction are unlikely to succeed in resource-limited countries , unless accompanied by measures for reducing dissemination of resistant isolates by improving sanitation . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [
"escherichia",
"coli",
"medicine",
"bacterial",
"diseases",
"infectious",
"diseases"
] | 2012 | Quinolone Resistance in Absence of Selective Pressure: The Experience of a Very Remote Community in the Amazon Forest |
The progressive failure of protein homeostasis is a hallmark of aging and a common feature in neurodegenerative disease . As the enzymes executing the final stages of autophagy , lysosomal proteases are key contributors to the maintenance of protein homeostasis with age . We previously reported that expression of granulin peptides , the cleavage products of the neurodegenerative disease protein progranulin , enhance the accumulation and toxicity of TAR DNA binding protein 43 ( TDP-43 ) in Caenorhabditis elegans ( C . elegans ) . In this study we show that C . elegans granulins are produced in an age- and stress-dependent manner . Granulins localize to the endolysosomal compartment where they impair lysosomal protease expression and activity . Consequently , protein homeostasis is disrupted , promoting the nuclear translocation of the lysosomal transcription factor HLH-30/TFEB , and prompting cells to activate a compensatory transcriptional program . The three C . elegans granulin peptides exhibited distinct but overlapping functional effects in our assays , which may be due to amino acid composition that results in distinct electrostatic and hydrophobicity profiles . Our results support a model in which granulin production modulates a critical transition between the normal , physiological regulation of protease activity and the impairment of lysosomal function that can occur with age and disease .
Aging and stress are thought to enhance neurodegenerative disease risk through the accumulation of misfolded and aggregated proteins [1–3] . The lysosome is the key degradative organelle within the cell [4] , and therefore plays a pivotal role in the maintenance of protein homeostasis . It contains specialized enzymes , called cathepsins , which work optimally at the acidic pH in this compartment and have a crucial role in processing and degrading proteins [5] . The transcription factor EB ( TFEB ) controls the expression of genes involved in lysosomal biogenesis and function [6 , 7] . TFEB dysregulation has been associated with neurodegenerative disease [8 , 9] and its overexpression may help to promote the clearance of protein aggregates [10 , 11] . Although , genetic and functional studies have implicated lysosomal dysfunction in the pathogenesis of multiple neurodegenerative diseases [12–14] , understanding of the molecular basis of this phenomenon remains incomplete . Heterozygous progranulin ( PGRN ) loss-of-function mutations lead to autosomal dominant transmission of the neurodegenerative disorder frontotemporal lobar degeneration ( FTLD ) with TAR DNA binding protein 43 ( TDP-43 ) inclusions [15–17] . The molecular function of the progranulin protein ( PGRN ) remained elusive until it was indelibly linked to lysosomal function by the finding that loss of both gene alleles results in the lysosomal storage disease , neuronal ceroid lipofuscinosis [18] . Progranulin localizes to lysosomes [19–22] where it may act to promote lysosomal biogenesis and function [20 , 23–25] . The progranulin ( PGRN ) protein can be proteolytically cleaved to liberate multiple cysteine-rich “granulin” peptides [26] . Granulins are highly conserved , disulfide-bonded miniproteins with unknown biological function [27–31] . Like progranulin , granulin peptides have been shown to localize to the endolysosomal compartment [32] , and can be generated through the action of cysteine proteases on progranulin [32–34] . Owing to the twelve cysteines and six disulfide bonds found in each cleaved granulin , these peptides adopt a stacked β-sheet configuration that is compact , structurally stable and potentially protease resistant [35] . Several lines of evidence exist that cleaved granulin peptides oppose the function of the full-length protein . While progranulin has proliferative [35 , 36] and anti-inflammatory [37 , 38] properties , granulin peptides have been shown to inhibit cell growth [35] and stimulate inflammation [38] . In addition , we have previously demonstrated a role for C . elegans granulins in selectively promoting the accumulation of TDP-43 , thereby exacerbating TDP-43 toxicity and potentially contributing to the pathogenesis of disease [39] . However , the mechanism by which granulins exert this specific regulation on TDP-43 metabolism remains unknown . C . elegans provides many advantages as a model system to study granulin function , including conservation of the progranulin gene , and the many available molecular and cell biology techniques . In this study , we further investigate the molecular mechanisms of C . elegans granulins on lysosomal function and protein homeostasis . We show that C . elegans granulins localize to the endolysosomal fraction . Granulin production increases with age and stress , and granulin expression reduces animal fitness by impairing lysosomal protease expression and activity . This prompts cells to activate a compensatory transcriptional program involving HLH-30/TFEB nuclear translocation and up-regulation of the transcription of HLH-30/TFEB-related genes . Overall , our findings highlight granulins as critical regulators of proteolytic lysosomal function and potential drivers of neurodegenerative disease pathogenesis .
We have previously shown that C . elegans progranulin ( pgrn-1 ) null mutants exhibit enhanced resistance to endoplasmic reticulum ( ER ) unfolded protein stress [40] . As a genetic null , pgrn-1 ( - ) animals produce neither full-length progranulin nor cleaved granulins; therefore , absence of either the holoprotein or the cleavage fragments could be responsible for the ER stress resistance . Based on our earlier finding that granulins could exacerbate TDP-43 toxicity [39] , we hypothesized that the bioactive granulins were responsible for inhibiting ER stress resistance . Hence , to isolate granulin activity , we expressed individual C . elegans granulins 1 , 2 and 3 at comparable levels in a pgrn-1 null background [39] . Granulin expression in a progranulin null background completely abolished the ER stress resistance phenotype ( Fig 1A ) . In contrast , animals over-expressing C . elegans full-length progranulin in a progranulin null background remained ER stress resistant ( Fig 1B ) . Over-expressed full-length progranulin was not cleaved under ER stress ( S1A–S1D Fig ) , and could promote ER stress resistance in the presence of granulin ( S1E Fig ) . Furthermore , transgenic expression of human tau protein and TDP-43 in a progranulin null background did not abrogate ER stress resistance ( S1F Fig ) . Taken together , these data suggest that it is the granulins , and not full-length progranulin , that specifically inhibit ER stress resistance . Given that granulins impair ER stress resistance , we wondered if they might more broadly impact protein homeostasis . Thus , we measured endogenous levels of heat shock protein HSP-4 , the nematode homolog of human BiP/Grp78 [41] . HSP-4/BiP expression is upregulated during the unfolded protein response ( UPR ) [42] . We found that granulin-expressing animals displayed a trend for increased basal expression of HSP-4/BiP on day 1 of adulthood , reaching significance in animals expressing granulin 2 and 3 ( Fig 1C ) . Therefore , in the absence of the progranulin holoprotein , granulin expression upregulates HSP-4 and this is indicative of UPR induction and perturbed protein homeostasis . While working with the granulin-expressing lines , we noted a decrease in overall animal fitness attributable to the granulins . Granulin production significantly reduced animal viability by lowering the number of eggs that hatched and slowing the development of animals to maturity ( Fig 1D ) . Granulin-expressing animals that did reach adulthood were smaller in size ( Fig 1E ) . Short-term associative learning can be assayed in C . elegans using a positive olfactory learning paradigm [43 , 44] . When granulin-expressing animals were tested in this assay they underperformed compared to controls ( Fig 1F ) , suggesting that granulin expression may result in neuronal dysfunction . These data , coupled with previous work by others on the function of progranulin [35–38] , suggest that granulins impair animal fitness , resistance to stress and neuronal function , while progranulin promotes these qualities . To establish the trafficking and localization of granulin peptides within a whole organism , we utilized microscopy and biochemistry techniques . First , we determined the sub-cellular localization of full-length progranulin using a translational progranulin reporter , PGRN-1::RFP , and organelle-specific markers . As expected , in cells that secrete progranulin , such as the intestine , the reporter co-localized with both a Golgi marker , mannosidase II ( Fig 2A ) , and a lysosomal marker , lysosomal-associated membrane protein 1 ( LMP-1 ) ( Fig 2B ) . However , in coelomocytes , a cell type that takes up but does not produce progranulin [45] , the progranulin reporter was only seen in the endolysosomal compartment ( Fig 2C–2E ) , suggesting that extracellular progranulin is transported through endosomes to reach the lysosome . Having established that progranulin can be trafficked from one tissue type to another , we next sought to better understand the subcellular localization of granulin peptides . To do so , we developed a protocol for subcellular fractionation of C . elegans . The purity of cytosolic , ER and endolysosomal fractions was confirmed with established markers ( S2 Fig ) . Individual granulins that were transgenically expressed also demonstrated lysosomal localization ( Fig 2F–2I ) . Therefore , C . elegans progranulin and granulins localize to the endolysosomal compartment . In C . elegans and mammals , progranulin production increases with age [45 , 46] and injury [47 , 48] . However , the degree to which granulin peptides are liberated has not been measured . We first asked if progranulin cleavage into granulins increases with age . Using our PGRN-1::RFP translational reporter , we found that granulin production does indeed increase in an age-dependent fashion ( Fig 3A and S3A and S3B Fig ) , suggesting that either an increase in expression and cleavage of progranulin , and/or an age-associated decline in granulin turnover , contributes to granulin accumulation . Granulin cleavage also increased in response to certain physiological stressors such as starvation ( Fig 3B and S3C and S3D Fig ) . Thus , age and stressful stimuli , such as starvation , appear to promote the cleavage of full-length progranulin into granulins . In order to determine the subcellular compartment in which cleaved granulin peptides are produced , we performed fractionation of fed or starved animals expressing the PGRN-1::RFP reporter . In fed animals , full-length progranulin was enriched in the endolysosomal fraction with very little lower molecular weight granulin observed in any fraction ( Fig 3C ) . Upon starvation , the cleaved granulins increased primarily in the endolysosomal fraction , confirming that the majority of the age and stress-induced granulins are , in fact , endolysosomal ( Fig 3C ) . Therefore , granulin peptides are produced in vivo in the endolysosomal compartment in a stress-responsive manner . Given that granulins impair organismal fitness , localize to the endolysosomal fraction and impair stress resistance , we next investigated their impact on lysosomal morphology . In C . elegans , coelomocytes scavenge and detoxify the pseudocoelomic cavity and therefore have a well-developed endo-lysosomal system [49] . Although we could not image coelomocyte lysosomes in granulin 1-expressing animals due to the presence of a GFP co-expression marker , we found that both loss of progranulin and expression of granulins 2 and 3 grossly deformed these organelles ( Fig 3D–3G ) . Lysosomes lost their spherical shape , more frequently exhibited membrane protrusions and tubular extensions ( Fig 3D–3G ) , and became smaller in size , reaching significance for pgrn-1 ( - ) animals and pgrn-1 ( - ) ; granulin 3 ( + ) animals ( Fig 3H ) . Together , these data suggest that granulin peptides accumulate in endolysosomes with age and starvation , where they , as well as loss of progranulin , may disrupt lysosomal morphology . As we observed that expressed granulins disrupt lysosomal morphology , we next assessed their effect on lysosomal function by measuring the expression level and enzymatic activity of lysosomal proteases in lysates from granulin-expressing C . elegans . Granulin expression resulted in decreased protein levels of ASP-3 , the nematode ortholog of mammalian cathepsin D ( CTSD ) , reaching significance in granulin 2-expressing animals ( Fig 4A ) . Expression of all granulins significantly reduced CPL-1 expression , the nematode ortholog of mammalian cathepsin L ( CTSL ) ( Fig 4B ) . This decrease in protease expression correlated with a decrease in protease activity ( Fig 4C and 4D ) , reaching significance in granulin 2 and 3-expressing animals for ASP-3 activity and granulin 1 and 2-expressing animals for CPR/CPL-1 activity . Overall , our data suggest that granulin peptides disrupt C . elegans lysosomal protease activity in vivo . As we observed differences between the three granulins in terms of the magnitude of their phenotypic effects within assays , we sought to determine whether these differences might be explained by variations in their amino acid sequence and physicochemical properties . C . elegans granulins 1 , 2 and 3 share less than 50% sequence identity among themselves ( Fig 5A ) , and less than 40% when excluding the highly conserved network of disulfide bonds . Electrostatic analysis ( Fig 5B ) shows that granulin 3 , located at the C-terminus of C . elegans PGRN-1 , is positively charged at neutral pH , while granulin 1 , the N-terminal granulin domain , remains negatively charged at all analyzed pH values ( pH = 4 to 8 ) . The central granulin 2 domain has little to no overall net peptide charge at neutral pH . A further comparison of granulin hydrophobicity ( Fig 5C ) shows that the central region of granulin 2 ( residues 202 to 221 ) and granulin 3 ( residues 309 to 326 ) is predominantly hydrophobic , as measured by Kyte and Doolittle ( K&D ) hydrophobicity scores greater than zero . In contrast , the K&D score for the corresponding region of granulin 1 ( res . 120 to 139 ) is slightly negative . While the functions of the individual C . elegans granulin domains remain to be further elucidated , these observed differences might suggest that each domain participates in unique protein-protein interactions ( PPIs ) , and thus differing roles in the endolysosomal system . We further compared the C . elegans granulin sequences with those of different species , including Homo sapiens ( H . sapiens ) , Mus musculus ( M . musculus ) and Danio rerio ( D . rerio ) ( Fig 5A ) . We found that C . elegans granulins share higher identity scores to certain granulins from other species than among themselves . Similar to C . elegans granulins , differences in pH-dependent electrostatics ( Fig 5B ) were noticeable for all species studied , with a recurring trend for the C-terminal granulin domains being the most positively charged . The low sequence identity and distinct physicochemical properties among the granulin domains were also observed for H . sapiens , M . musculus and D . rerio , contrasting with the highly conserved network of disulfide bonds . Taken together , these data highlight the importance of the amino acid residues situated outside of the well-conserved granulin sequence consensus for contributing to the charge and hydrophobicity profiles of each granulin domain . These may drive unique recognition patterns for PPIs that may ultimately be relevant in a disease context . To determine if granulin-induced disruption of lysosomal morphology and function promoted a transcriptional response , we performed RNA-seq profiling of wild-type , pgrn-1 ( - ) and granulin-expressing animals ( S1–S5 Tables ) . Visual inspection of the RNA sequencing reads confirmed a high and comparable expression of granulin 1 , 2 and 3 transgenes , as well as a read drop-out in progranulin null animals ( S4A Fig ) . Wild-type animals had a low but detectable expression of endogenous progranulin transcript ( S4A Fig ) . We first compared pgrn-1 ( - ) or pgrn-1 ( - ) ; granulin animals to wild-type animals . Compared to wildtype , a total of 7084 differentially expressed genes ( DEGs ) were identified across all strains ( Fig 6A and S4B Fig ) . The majority of DEGs identified for pgrn-1 ( - ) animals were down-regulated compared to wild-type animals . These DEGs were enriched for GO terms associated with growth , development , cation and sugar binding ( S4C and S4D Fig ) . In contrast , the majority of DEGs for granulin-expressing animals were up-regulated compared to both wild-type and pgrn-1 ( - ) animals ( Fig 6A and S4B Fig ) . GO term analysis for DEGs in granulin-expressing animals showed a shared enrichment in genes associated with lysosomal function , including protein metabolic process and hydrolase activity acting on ester bonds ( S4E–S4K and S5 Figs ) . Expression of granulin 2 resulted in the highest number of DEGs compared to both wild-type and pgrn-1 ( - ) animals , followed by granulin 3 and then granulin 1 ( Fig 6A and S4B Fig ) . The observed overlap in enriched GO terms on granulin 2 and 3 expression further suggests similarities between these two granulins compared to granulin 1 , and also reflects the phenotype severity observed in development and behavioral assays . Interestingly , the upregulated DEGs identified in pgrn-1 ( - ) ; granulin 3 ( + ) animals were significantly enriched for genes whose promoters contained the putative TFEB binding site E-box sequence 5’-CACGTG-3’ ( P = 0 . 011 ) . This trend was also observed in the upregulated DEGs for pgrn-1 ( - ) ; granulin 1 ( + ) ( P = 0 . 149 ) and pgrn-1 ( - ) ; granulin 2 ( + ) ( P = 0 . 097 ) but did not reach statistical significance . TFEB is the master lysosomal transcription factor that regulates lysosomal biogenesis and autophagy [6 , 7] , and the C . elegans TFEB is HLH-30 [50] . In response to starvation , stressful stimuli and aging , HLH-30/TFEB translocates from the cytosol to the nucleus to activate its transcriptional targets [6 , 7 , 50 , 51] . This program , known as the Coordinated Lysosomal Expression and Regulation ( CLEAR ) response induces expression of genes involved in lysosomal function and autophagy , including progranulin . We assessed HLH-30/TFEB cytoplasmic versus nuclear localization in control , pgrn-1 ( - ) and granulin expressing animals . Granulin expression promoted nuclear localization of HLH-30/TFEB , reaching significance in granulin 3-expressing animals ( Fig 6B and 6C ) . This effect was not seen in pgrn-1 ( - ) animals where a much lower number of DEGs were identified , and was also not observed in pgrn-1 ( - ) animals expressing human tau or TDP-43 protein ( S6A Fig ) . These results suggest that the disruption of lysosomal morphology and protein homeostasis seen in granulin-expressing animals leads to a specific compensatory translocation of HLH-30/TFEB from the cytosol to the nucleus . When granulin-expressing animals were crossed into a wildtype background , the presence of wildtype progranulin partially mitigated the negative effects of granulin-expression on development ( S6B Fig ) , lysosome morphology ( S6C and S6D Fig ) and HLH-30/TFEB localization ( S6E Fig ) . Interestingly , granulin-expression in a wildtype background resulted in higher ER stress sensitivity than granulin-expression in a progranulin null background ( S6F Fig ) . We speculate that ER stress may promote the cleavage of endogenous PGRN , resulting in even higher levels of cleaved granulins ( endogenous and transgenic granulins ) and enhanced ER stress sensitivity . These data further suggest a reciprocal relationship between full-length progranulin and cleaved granulins , and highlights that their relative levels may be important for normal animal development and fitness . To determine if the upregulation of TFEB target genes was a compensatory transcriptional response in granulin-expressing animals , we crossed these animals into an hlh-30 ( - ) null background . When lacking hlh-30 , granulin-expressing animals had further impairments in overall fitness , with fewer growing to adulthood ( Fig 6D ) and more arresting at early larval stages ( Fig 6E ) . Together , these data demonstrate that granulin expression , even in the absence of stress or starvation , is sufficient to activate a compensatory CLEAR response and induce expression of genes containing TFEB binding sites . Overall , the ability of granulins to 1 ) impair a proteotoxic stress response , 2 ) disrupt lysosomal morphology , 3 ) direct TFEB to the nucleus and 4 ) induce a CLEAR response indicates that granulin-dependent impairment of lysosomal function negatively impacts cellular protein homeostasis ( Fig 6F ) .
We have previously shown in C . elegans that expression of granulin peptides enhances TDP-43 toxicity and prevents its degradation [39] . In this study , we sought to understand the mechanism by which granulins exert their effects and determine if they more broadly impacted protein homeostasis . We found that granulins are produced in an age and stress-dependent manner , and consequently impair lysosomal protease expression and activity . Their expression negatively impacts cellular protein homeostasis and drives a compensatory lysosomal stress response in an attempt to up-regulate HLH-30/TFEB-regulated genes . These effects manifest as an overall decrease in animal fitness . This study contributes a new dimension to our understanding of the regulation of lysosomal proteostasis via the identification of C . elegans granulins as age and stress-produced peptides that impair overall animal fitness by reducing lysosomal function . C . elegans granulins , similar to the human peptides , localize to the endolysosomal compartment [32] . Granulins are composed of evolutionarily conserved stacked beta hairpins stabilized by disulfide bonds , which are often found in natural protease inhibitors [52] . This highly compact and stable structure is thought to confer resistance to denaturation and protection against proteolytic cleavage in the lysosomal environment [53] . Indeed , a role for granulins in regulating protease maturation has previously been demonstrated in plant cysteine proteases that incorporate a granulin domain C-terminal to the catalytic domain , such as RD21 in A . thaliana [54] . In further support of granulins as regulators of protease activity , homozygous progranulin mutation carriers develop a progressive myoclonic epilepsy syndrome that phenocopies loss of function mutations in another lysosomal protease inhibitor , cystatin B [18 , 55] . Recent studies have shown that human full-length progranulin and individual granulin domains may physically interact with CTSD and stimulate the enzymatic activity of the protease [25 , 56–58] . However , in the absence of full-length protein , C . elegans granulins promote a distinct phenotype of impaired resistance to ER stress , delayed growth , decreased CTSD and CTSB/L activity and activation of the CLEAR transcriptional program . Granulins likely play a normal physiological role in regulating protease expression and activity . Given their ability to promote the CLEAR program , granulins may serve as a signal for stress or impaired health that requires regulated checks on protease activity , perhaps to limit inflammation . This would be consistent with the role of progranulin in complement-mediated synaptic pruning by microglia [59] . We speculate that under conditions of progranulin haploinsufficiency , the normal balance between progranulin and granulins becomes skewed towards excessive granulins . In excess , the inhibitory effect of granulins upon protease activity impairs the function of lysosomes; with age , the natural compensatory mechanisms such as the CLEAR program become overwhelmed , resulting in cellular dysfunction . When this occurs in neurons and/or support cells such as microglia , the end result may be neurodegeneration . Because granulins increase with age , it remains possible that accumulation of granulins directly contribute to the proteostatic pressures associated with increasing age . Comprehensive measures of progranulin-to-granulin ratios with age and in progranulin mutation carriers are needed . The lentiviral delivery of progranulin to degenerating brain regions protects against neurotoxicity and cognitive defects in mouse models of Parkinson’s disease [60] and Alzheimer’s disease [61] . As such , efforts to increase progranulin production in patients are underway [62–65] . However , a more recent study has suggested that progranulin delivery to brain promotes in T-cell infiltration and neuronal and glial degeneration [66] . Progranulin cleavage and granulin levels were not measured in these studies , and may account for differences in the observed results . Progranulin is a highly conserved protein [27 , 29 , 30] . The number of granulin domains has increased through phylogeny from one in Dictyostelium discoideum and plants , three in nematodes to seven-and-a-half in humans [29 , 54] . It is intriguing to speculate that this expansion in cleavage fragments could lead to regulation of additional proteases . In support of this , we find that the amino acid residues situated outside of the well-conserved granulin sequence consensus contribute to distinct charge and hydrophobic profiles for each granulin domain . These unique characteristics may be important for driving specific protein-protein interactions and thus different roles in the cellular environment . Indeed , the distinct effects of granulin 2 and 3 on protein homeostasis , lysosomal function and TDP-43 toxicity [39] , as compared to granulin 1 , may suggest functional differences between granulins . Our results establish age-regulated granulins as modulators of lysosomal function , and suggest that a toxic gain of granulin function , rather than or in addition to simply loss of full-length progranulin , may contribute to FTLD disease pathogenesis . This could explain why progranulin loss-of-function mutations are transmitted in an autosomal dominant fashion . The presence of granulins only in the haploinsufficiency state could explain why TDP-43 pathology is not seen in the null state [18] . Several lysosomal proteases that cleave progranulin have recently been identified [32–34] , although how those proteases decide when and where to cleave progranulin remains unknown . This study prompts several important follow up questions regarding the rate and order in which granulins are liberated from progranulin , how pH changes impact the predicted association of granulins with lysosomal proteases and whether increased granulin impact other neurodegenerative disorders such as Alzheimer’s disease . The current study also has implications for therapeutic progranulin repletion efforts , as care should be taken to determine whether replacement progranulin is processed into granulins . Finally , our findings suggest that in addition to progranulin repletion , prevention of progranulin cleavage into granulins could represent a rational therapeutic target in neurodegeneration .
C . elegans strains were cultured at 20 °C according to standard procedures [67] . Some strains were provided by the Mitani Laboratory ( National Bioresource Project , Japan ) at the Tokyo Women’s Medical University and the Caenorhabditis Genetics Center ( CGC ) at the University of Minnesota . Strain descriptions are at www . wormbase . org . The N2E control strain was used as the wild-type strain . The pgrn-1 ( tm985 ) strain has a 347 bp deletion in the pgrn-1 gene resulting in a null allele [45] . The following C . elegans strains were used in this study: CF3050 pgrn-1 ( tm985 ) I AWK33 pgrn-1 ( tm985 ) I; rocIs1[Ppgrn-1+SignalSequence::granulin1::FLAG::polycistronic mCherry + Punc-122::GFP] AWK43 pgrn-1 ( tm985 ) I; rocEx14[Ppgrn-1+SignalSequence::granulin2::FLAG::polycistronic mCherry + Pmyo-2::GFP] AWK107 pgrn-1 ( tm985 ) I; rocIs5[Ppgrn-1+SignalSequence::granulin3::FLAG::polycistronic mCherry + Pmyo-2::GFP] AWK308 N2E; rocIs1[Ppgrn-1+SignalSequence::granulin1::FLAG::polycistronic mCherry + Punc-122::GFP] AWK309 N2E; rocEx14[Ppgrn-1+SignalSequence::granulin2::FLAG::polycistronic mCherry + Pmyo-2::GFP] AWK310 N2E; rocIs5[Ppgrn-1+SignalSequence::granulin3::FLAG::polycistronic mCherry + Pmyo-2::GFP] AWK459 pgrn-1 ( tm985 ) I; muIs216[Paex-3::huMAPT 4R1N +Pmyo-3::RFP] CF3588 pgrn-1 ( tm985 ) I; muIs206[Pegl-3::TDP-43::GFP] AWK524 pgrn-1 ( tm985 ) I; muIs189[Ppgrn-1::pgrn-1::polycishronic mCherry +Podr-1::CFP] AWK466 pgrn-1 ( tm985 ) I; muIs189[Ppgrn-1::pgrn-1::polycishronic mCherry +Podr-1::CFP]; rocEx14[Ppgrn-1+SignalSequence::granulin2::FLAG::polycistronic mCherry + Pmyo-2::GFP] CF3778 pgrn-1 ( tm985 ) I; muIs213[Ppgrn-1::pgrn-1::RFP] AWK181 pgrn-1 ( tm985 ) I; unc-119 ( ed3 ) III; pwIs503[vha6p::mans::GFP + Cb unc-119 ( + ) ]; muIs213[Ppgrn-1::pgrn-1::RFP] AWK360 pgrn-1 ( tm985 ) I; unc-119 ( ed3 ) III; pwIs50[Plmp-1::lmp-1::GFP + Cbr-unc-119 ( + ) ]; muIs213[Ppgrn-1::pgrn-1::RFP] AWK395 pgrn-1 ( tm985 ) I; unc-119 ( ed3 ) III; cdIs54[pcc1::MANS::GFP + unc-119 ( + ) + myo-2::GFP]; muIs213[Ppgrn-1::pgrn-1::RFP] AWK374 pgrn-1 ( tm985 ) I; bIs34[rme-8::GFP + rol-6 ( su1006 ) ]; muIs213[Ppgrn-1::pgrn-1::RFP] MAH235 sqIs19[Phlh-30::hlh-30::gfp + rol-6 ( su1006 ) ] AWK403 pgrn-1 ( tm985 ) I; sqIs19[Phlh-30::hlh-30::gfp + rol-6 ( su1006 ) ] AWK404 pgrn-1 ( tm985 ) I; sqIs19[Phlh-30::hlh-30::gfp + rol-6 ( su1006 ) ]; rocIs1[Ppgrn-1+SS::granulin1::FLAG::polycistronic mCherry] AWK405 pgrn-1 ( tm985 ) I; sqIs19[Phlh-30::hlh-30::gfp + rol-6 ( su1006 ) ]; rocEx14[Ppgrn-1+SS::granulin2::FLAG::polycistronic mCherry + Pmyo-2::GFP] AWK406 pgrn-1 ( tm985 ) I; sqIs19[Phlh-30::hlh-30::gfp + rol-6 ( su1006 ) ]; rocIs5[Ppgrn-1+SS::granulin3::FLAG::polycistronic mCherry + Pmyo-2::GFP] AWK467 N2E; sqIs19[Phlh-30::hlh-30::gfp + rol-6 ( su1006 ) ]; rocIs1[Ppgrn-1+SS::granulin1::FLAG::polycistronic mCherry] AWK469 N2E; sqIs19[Phlh-30::hlh-30::gfp + rol-6 ( su1006 ) ]; rocEx14[Ppgrn-1+SS::granulin2::FLAG::polycistronic mCherry + Pmyo-2::GFP] AWK471 N2E; sqIs19[Phlh-30::hlh-30::gfp + rol-6 ( su1006 ) ]; rocIs5[Ppgrn-1+SS::granulin3::FLAG::polycistronic mCherry + Pmyo-2::GFP] AWK546 pgrn-1 ( tm985 ) I; sqIs19[Phlh-30::hlh-30::gfp + rol-6 ( su1006 ) ]; muIs216[Paex-3::huMAPT 4R1N +Pmyo-3::RFP] AWK547 pgrn-1 ( tm985 ) I; sqIs19[Phlh-30::hlh-30::gfp + rol-6 ( su1006 ) ]; muIs206[Pegl-3::TDP-43::GFP] JIN1375 hlh-30 ( tm1978 ) IV AWK514 pgrn-1 ( tm985 ) I; hlh-30 ( tm1978 ) IV AWK516 pgrn-1 ( tm985 ) I; hlh-30 ( tm1978 ) IV; rocIs1[Ppgrn-1+SS::granulin1::FLAG::polycis mCherry] AWK518 pgrn-1 ( tm985 ) I; hlh-30 ( tm1978 ) IV; rocEx14 [Ppgrn-1+SS::granulin2::FLAG::polycistronic mCherry + Pmyo-2::GFP] AWK519 pgrn-1 ( tm985 ) I; hlh-30 ( tm1978 ) IV; rocIs5 [Ppgrn-1+SS::granulin3::FLAG::polycistronic mCherry + Pmyo-2::GFP] AWK521 pgrn-1 ( tm985 ) I; hlh-30 ( tm1978 ) IV; muIs189[Ppgrn-1::pgrn-1::polycistronic mCherry +Podr-1::CFP] AWK296 N2E; Ex[Pced-1::asp-3::mrfp + pRF4 ( rol-6 ) ]; unc-119 ( ed3 ) III; pwIs50[Plmp-1::lmp-1::GFP + Cbr-unc-119 ( + ) ] AWK333 pgrn-1 ( tm985 ) I; Ex[Pced-1::asp-3::mrfp + pRF4 ( rol-6 ) ]; unc-119 ( ed3 ) III; pwIs50[Plmp-1::lmp-1::GFP + Cbr-unc-119 ( + ) ] AWK247 pgrn-1 ( tm985 ) I; pwls50[lmp-1::GFP + Cbr-unc-119 ( + ) ];rocEx14 [Ppgrn-1+SS::granulin2::FLAG::polycis tronic mCherry + Pmyo-2::GFP] AWK334 pgrn-1 ( tm985 ) I; Ex[Pced-1::asp-3::mrfp + pRF4 ( rol-6 ) ]; unc-119 ( ed3 ) III; pwIs50[Plmp-1::lmp-1::GFP + Cbr-unc-119 ( + ) ]; rocIs5[Ppgrn-1+SS::granulin3::FLAG::polycis tronic mCherry + Pmyo-2::GFP] AWK177 asp-3 ( tm4450 ) X VM487 nmr-1 ( ak4 ) II To generate strains expressing individual granulins , each granulin was amplified separately from wild-type C . elegans progranulin cDNA as previously described [39] . ER stress assays were performed as previously described [40] . L4 stage animals were allowed to lay eggs overnight . Fifty synchronized eggs were transferred to seeded plates . After three days , the fraction of animals that developed to the L4 stage was quantified . L4 animals were staged , grown at 20 °C overnight and imaged the following day as day 1 adults . Animals were mounted on a 2% agarose pad with 25 mM sodium azide ( Spectrum Chemical , #SO110 ) and imaged using a Zeiss AxioImager microscope at 10 x . Body size was measured in ImageJ software using the skeletonize function . Short-term associative learning assays were performed as previously described [43 , 44] . Sixty L4 stage animals were allowed to lay eggs overnight ( ~sixteen hours ) . Adult worms and hatched larvae were washed off the plates with M9 buffer . Eggs were collected with a cell scraper and transferred to a newly seeded plate by chunking . These eggs were allowed to develop to early L4 stage and 200 μl of 20 mM FUDR ( Fisher Scientific , #AC227601000 ) was added to prevent development of progeny and overgrowth of plates . At each time point , animals were collected from plates with ice cold M9 and washed once to remove food . The worm pellet was resuspended 1:1 in freshly made ice cold RIPA buffer ( 50 mM Tris pH 7 . 4 , 150 mM NaCl , 5 mM EDTA , 0 . 5% SDS , 0 . 5% SDO , 1% NP-40 , 1 mM PMSF , cOmplete protease inhibitor ( Roche , #04693124001 ) and PhosSTOP phosphatase inhibitor ( Roche , #04906837001 ) , 0 . 3 mM Pefabloc ( Roche , #11429868001 ) ) . Worms were transferred to Eppendorf tubes and sonicated for 4 cycles of 1 minute on and 2 minutes off ( BioRuptor , Diagenode ) . Lysates were centrifuge for 5 minutes at 13 , 000 rpm at 4 °C . Supernatant was transferred to a fresh Eppendorf tube and samples were boiled at 95 °C ( with 4x LDS , 10% reducing agent ) for 5 minutes and analyzed by SDS PAGE . 10–50 μg total protein was resolved on 4–12% gradient SDS-PAGE gels and transferred to PVDF . Commercial antibodies used for Western blotting were the following: Anti-HSP-4/BiP ( Novus Biologicals , #NBP1-06274 , 1:1000 dilution ) Anti-RFP ( GenScript , #A00682 , 1:1000 dilution ) Anti-FLAG ( Sigma , #F3165 , 1:1000 dilution ) Anti-LMP-1 ( Developmental Studies Hybridoma Bank , #LMP1 , 1:100 dilution ) Anti-HSP-70/HSC-70 ( Santa Cruz Biotechnology Inc . , #sc-33575 , 1:1000 dilution ) Anti-calnexin ( Novus Biologicals , #NBP1-97476 , 1:1000 dilution ) Anti-CPL-1 ( Abcam , #ab58991 , 1:500 dilution ) Anti-actin ( EMD Millipore , #MAB1501R , 1:5000 dilution ) Goat anti-mouse ( LI-COR IRDye 800CW , #925–32210 , 1:10 , 000 dilution ) Goat anti-rabbit ( LI-COR IRDye 800CW , #925–32211 , 1:10 , 000 dilution ) Donkey anti-goat ( LI-COR IRDye 800CW , #925–32214 , 1:10 , 000 dilution ) Donkey anti-mouse ( LI-COR IRDye 680RD , #925–68072 , 1:10 , 000 dilution ) Antibodies made in-house and used for Western blotting were the following: Anti-granulin 1 ( RB2481 , Biomatik , epitope HQCDAETEC ( acm ) SDDET , 1:1000 dilution ) Anti-granulin 3 ( RB2487 , Biomatik , epitope CTVLMVESARSTLKL , 1:1000 dilution ) Anti-ASP-3 ( Fred Hutchinson , epitope CTGPTDVIKKIQHKIG , 1:1000 dilution ) Imaging and quantification were performed on the LI-COR Odyssey Infrared System . Three independent blots were performed . Animals were mounted on microscope slides with 2% agarose pads containing 30 mM levamisole hydrochloride ( Fisher Scientific , #AC187870100 ) and imaged using a Zeiss LSM 700 laser-scanning confocal microscope using 488 nm and 561 nm lasers and 63x and 100x objectives . L1 animals were imaged 1–2 h after hatching . Z-stacks were taken every 0 . 7 μm . Image processing was carried out using ImageJ software . A maximum intensity projection of the z-stack for each animal was created . Images at 488 nm and 561 nm were overlaid and analyzed for co-localization . Thirty L4 stage animals were picked to 60 x 10 cm plates per strain . Plates were confluent with mixed stage animals after four days growth at 20 °C . Progranulin cleavage was observed after starving animals for an additional seventy-two hours at 20 °C . A lysosomal fraction was isolated from a light mitochondrial-lysosomal fraction as previously described [68] with the following modifications . Animals were collected in 0 . 25 M sucrose ( pH 7 . 2 ) and washed twice with 0 . 25 M sucrose . Lysosomes and mitochondria were separated using a discontinuous Nycodenz ( Progen Biotechnik , Germany , #1002424 ) density gradient . Lysosomes were collected from the 19 . 8% / sucrose interface and the 26 . 3 / 19 . 8% interface and pooled . Lysosomes were diluted five times with 0 . 25 M sucrose , and pelleted at 37 , 000 × g for 15 minutes . Cytosolic , ER and lysosomal fractions were confirmed by immunoblotting for specific subcellular fraction markers ( LAMP-1 , HSC-70 , calnexin ) . Protease activity was measured using commercially available kits ( BioVision Cathepsin D Activity Fluorometric Assay Kit , #K143-100 and BioVision Cathepsin L Activity Fluorometric Assay Kit , #K142-100 ) . Animals were staged as for immunoblotting , but without the addition of 20 mM FUDR . At day 1 of adulthood , worms were collected from plates with ice cold M9 and washed twice to remove food . Worm pellets were resuspended in 1% NP-40 buffer ( Fisher Scientific ) without protease inhibitors and frozen at -80 °C overnight . Pellets were thawed and sonicated for 4 cycles of 1 min on and 2 min off ( BioRuptor , Diagenode ) . Lysates were centrifuged for 5 minutes at 13 , 000 rpm at 4 °C and supernatant was transferred to a fresh tube . 0 . 25 μg total protein per sample was used per assay and samples from one strain were run in triplicate . Fluorescence measurements were taken every minute at 25 °C ( Infinite M200 , Tecan ) . As controls , 250 nM Pepstatin A ( for pan-aspartyl protease inhibition in CTSD assay , BioVision ) or 10 μM CA-074 ( for Cathepsin B inhibition , EMD Millipore , #205530 ) and 10 μM CTSLiII ( for Cathepsin L inhibition , EMD Millipore , #219426 ) were added to the lysate and pre-incubated for 10 minutes on the bench at room temperature . Linear regression was performed on at least 30 minutes of data to calculate the rate of enzyme activity . Sequences for C . elegans ( Q9U362 ) , Homo sapiens ( P28799 ) and Mus musculus ( P28798 ) PGRN were extracted from Uniprot ( The Uniprot Consortium , 2019 ) , while Danio rerio PGRNb ( AAH96854 . 1 ) sequence was obtained from National Center for Biotechnology Information ( NCBI ) Protein database ( https://www . ncbi . nlm . nih . gov ) . Amino acid multiple sequence alignment was performed using the MAFFT online service ( version 7 , https://mafft . cbrc . jp/alignment/server/ ) [69] . The EMBOSS Needle server was used for pairwise sequence alignment between C . elegans granulin 1 , granulin 2 and granulin 3 and individual granulin domains from H . sapiens , M . musculus and D . rerio ( https://www . ebi . ac . uk/Tools/psa/emboss_needle/ ) [70] . Identification of granulin domains from the full-length sequences was based on sequence similarity to H . sapiens granulin A using the Basic Local Alignment Search Tool protein ( BLASTp ) server ( https://blast . ncbi . nlm . nih . gov/Blast . cgi ) . Granulin A ( PDB ID: 2JEY . A ) was used as a reference for homology modeling of all granulin domains [71] using the Prime software . Electrostatic analysis ranging from pH 4 to 8 was performed on the in silico models with the software propKa 3 . 1 [72] . Kyte & Doolitle ( K&D ) hydrophobicity scales were obtained from the ExPASy Bioinformatics Resource Portal ( https://web . expasy . org ) for PGRN sequence of all species here studied . For the K&D per-residue score , a window size of 5 was used , i . e . the final score for a given residue i is the sum of the scale values for i and i-2 , i-1 , i+1 and i+2 . Total RNA was isolated from wild-type ( N2E ) , pgrn-1 ( - ) , pgrn-1 ( - ) ; granulin 1 ( + ) , pgrn-1 ( - ) ; granulin 2 ( + ) and pgrn-1 ( - ) ; granulin 3 ( + ) expressing animals synchronized at day 1 of adulthood . Animals were collected from plates with ice cold M9 and washed three times to remove OP50 food . After harvesting , the animals were resuspended in QIAzol ( Qiagen #79306 ) and flash frozen in liquid nitrogen . RNA was extracted and purified using a Qiagen miRNeasy kit ( Qiagen #217004 ) . Samples were extracted in quadruplicate ( four biological replicates for each strain ) , for a total of 20 samples . Total RNA was quantified using the RiboGreen assay ( ThermoFisher , #R11490 ) and RNA quality was checked using an Agilent TapeStation 4200 ( Agilent ) . RNA Integrity Numbers ( eRINs ) were >8 in all the samples . Libraries for RNA-seq were prepared using the Illumina TruSeq library preparation protocol ( Illumina Inc ) , multiplexed into a single pool and sequenced using an Illumina HiSeq 4000 sequencer across 4 PE 2 x 75 lanes on a single flowcell . After demultiplexing , we obtained between 13 and 32 million reads per sample , each one 75 paired end bases long . Quality control was performed on base qualities and nucleotide composition of sequences . Alignment to the C . elegans genome ( ce11 ) was performed using the STAR spliced read aligner [73] with default parameters . Additional QC was performed after the alignment to examine the following: level of mismatch rate , mapping rate to the whole genome , repeats , chromosomes , and key transcriptomic regions ( exons , introns , UTRs , genes ) . Between 92 and 93% of the reads mapped uniquely to the worm genome . Total counts of read fragments aligned to candidate gene regions within the C . elegans reference gene annotation were derived using HTS-seq program and used as a basis for the quantification of gene expression . Only uniquely mapped reads were used for subsequent analyses . Following alignment and read quantification , we performed quality control using a variety of indices , including sample clustering , consistency of replicates , and average gene coverage . One sample for pgrn-1 ( - ) ; granulin 1 ( + ) was excluded from analysis as a quality control outlier . Differential expression analysis was performed using two parallel approaches , the EdgeR Bioconductor package [74] , and voom [75] . Differentially expressed genes ( DEGs ) were selected based on False Discovery Rate ( FDR , Benjamini-Hochberg adjusted p-values ) estimated at ≤ 5% . There was a large overlap between DEGs identified by edgeR and voom ( edgeR: 89 . 0% common DEGs with voom ( 6307/7084 ) , voom: 93 . 9% common DEGs with edgeR ( 6307/6714 ) ) . Clustering and overlap analyses were performed using the Bioconductor packages within the statistical environment R ( www . rproject . org/ ) . Gene Ontology annotation was performed using DAVID ( david . abcc . ncifcrf . gov/ ) and GOrilla [76 , 77] . The promoter regions of all differentially regulated transcripts were analyzed for the presence of the C . elegans TFEB/HLH-30 binding site E-box sequence 5’-CACGTG-3’ . Enrichment of TFEB binding sites was tested by comparison to the expected distribution based on 10 , 000 random permutations . A permutation test was used to calculate p-values . Forty L4 animals were picked , grown at 20 °C overnight and imaged the following day as day 1 adults . The nuclear localization of HLH-30::GFP was imaged using a Zeiss AxioImager microscope at 10x . Animals were imaged within 5 minutes of mounting on a 2% agarose pad with 25mM sodium azide ( Spectrum Chemical , #SO110 ) . Data from three independent experiments were pooled . | Progressive decline in maintenance of protein homeostasis clearly contributes to the development of neurodegenerative disorders , yet the molecular basis of this decline is poorly understood . Here , we take advantage of molecular genetic techniques available in the model organism C . elegans to investigate the mechanism underlying neurodegenerative disease due to mutations in the progranulin gene . We find that age , gene mutation and physiological stress lead to the accumulation of lysosomal granulins ( the cleavage products of the progranulin protein ) thereby disrupting cellular protein homeostasis . Granulin expression impairs animal fitness , resistance to stress and neuronal function , and stimulates a lysosomal stress response in an attempt to up-regulate lysosomal genes and restore normal function . Our findings are particularly important because they suggest a new , rational target—inhibition of progranulin cleavage into granulins—for neurodegenerative disease therapy . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"invertebrates",
"lysosomes",
"medicine",
"and",
"health",
"sciences",
"caenorhabditis",
"enzymes",
"endoplasmic",
"reticulum",
"cell",
"processes",
"enzymology",
"cell",
"disruption",
"animals",
"animal",
"models",
"physiological",
"processes",
"caenorhabditis",
"elegans",
"model",
"organisms",
"homeostasis",
"experimental",
"organism",
"systems",
"cellular",
"structures",
"and",
"organelles",
"research",
"and",
"analysis",
"methods",
"sequence",
"analysis",
"specimen",
"preparation",
"and",
"treatment",
"sequence",
"alignment",
"animal",
"studies",
"proteins",
"bioinformatics",
"gene",
"expression",
"biochemistry",
"mechanical",
"treatment",
"of",
"specimens",
"eukaryota",
"specimen",
"disruption",
"cell",
"biology",
"secretory",
"pathway",
"physiology",
"database",
"and",
"informatics",
"methods",
"genetics",
"nematoda",
"biology",
"and",
"life",
"sciences",
"proteases",
"organisms"
] | 2019 | Age- and stress-associated C. elegans granulins impair lysosomal function and induce a compensatory HLH-30/TFEB transcriptional response |
Anisakiasis is an emerging public health problem , caused by Anisakis spp . nematode larvae . Anisakiasis presents as variable and unspecific gastrointestinal and/or allergic clinical symptoms , which accounts for the high rate of misdiagnosed cases . The aim of this study was to characterize the early cellular ( 6–72 h p . i . ) and molecular ( 6 h p . i . ) immune response and general underlying regulatory mechanism in Anisakis infected rats . Each Sprague-Dawley rat was infected with 10 Anisakis spp . larvae by gastric intubation . Tissues with visible lesions were processed for: i ) classic histopathology ( HE ) , immunofluorescence ( CD3 , iNOS , S100A8/A9 ) , and transmission electron microscopy ( TEM ) ; ii ) target genes ( Il1b , Il6 , Il18 , Ccl3 , Icam1 , Mmp9 ) and microRNA ( Rat Immunopathology MIRN-104ZF plate , Quiagen ) expression analysis; and iii ) global DNA methylation . Histopathology revealed that Anisakis larval migration caused moderate to extensive hemorrhages in submucosal and epimysial/perimysial connective tissue . In stomach and muscle , moderate to abundant mixed inflammatory infiltrate was present , dominated by neutrophils and macrophages , while only mild infiltration was seen in intestine . Lesions were characterized by the presence of CD3+ , iNOS+ , and S100A8/A9+ cells . The greatest number of iNOS+ and S100A8/A9+ cells was seen in muscle . Il6 , Il1b , and Ccl3 showed particularly strong expression in stomach and visceral adipose tissues , but the order of expression differed between tissues . In total , three miRNAs were differentially expressed , two in stomach ( miRNA-451 and miRNA-223 ) and two in intestine ( miRNA-451 and miRNA-672 ) . No changes in global DNA methylation were observed in infected tissues relative to controls . Anisakis infection induces strong immune responses in infected rats with marked induction of specific proinflammatory cytokines and miRNA expression . Deciphering the functional role of these cytokines and miRNAs will help in understanding the anisakiasis pathology and controversies surrounding Anisakis infection in humans .
Foodborne parasitic diseases have received increasing attention in recent decades . Some are considered emerging public health issues , due to changes in various factors influencing the propagation of parasites that cause them [1] . This is especially valid for several helminthiases , which have seen a rise in reported cases outside the natural range of their causative agents , mostly due to changes in dietary habits , such as eating raw or undercooked food , and increased demand and availability of “exotic” food [1 , 2] . The Foodborne Disease Burden Epidemiology Reference Group ( FERG ) published a list of parasites that could be transmitted to humans by food . The list includes several helminth taxa considered to present substantial disease burden , such as Fasciola spp . , Trichinella spp . , Echinococcus spp . , Clonorchis spp . , Opistorchis spp . , Taenia solium , Anisakis simplex , and Ascaris lumbricoides [3] . Although A . simplex was subsequently removed from the priority list as being an uncommon foodborne parasite [4] , the European Food Safety Authority ( EFSA ) has listed anisakid nematodes as parasites of high public health importance [5] . In addition , a recent study ranked Anisakidae within the top 10 and top 5 foodborne parasites important in Northern and South-Western Europe regions , respectively [6] . This is also evidenced from the recently published analysis of a 10-year hospital discharge records that reported 370 confirmed cases of anisakiasis in Italy , with an annual incidence of 1 . 5 to 15 . 5 per 100000 hospitalizations [7] . Conversely , only 37 cases of anisakiasis were estimated over a 5-year period in France based on national survey among hospital Parasitology laboratories in France[8] . However , a recent risk assessment analysis indicated a substantial underestimation of Anisakis infection in European countries , predicting 7700–8320 annual cases for Spain only [9] . Members of the genus Anisakis , ( with currently nine valid species ) are parasitic nematodes with an indirect life cycle that utilize marine mammals , primarily cetaceans , as definitive hosts , planktonic crustaceans as first intermediate hosts , and fish and cephalopods as paratenic hosts [10] . In definitive hosts , these parasites induce chronic changes in the gastrointestinal mucosa ranging from superficial erosions to deep ulcerations . Histopathologically , lesions are characterized by lymphoplasmacytic gastritis , granulomatous or eosinophilic inflammation with the presence of multinucleated giant cells , hemosiderosis , and fibrosis [11–14] . Following ingestion of raw or undercooked seafood infected with live Anisakis third-stage larvae ( L3 ) , humans can become accidental hosts; thus contracting a disease termed anisakiasis [2 , 15] . In humans , Anisakis larvae cannot reach adult stage , and although infection can be asymptomatic , it is often followed by an acute clinical state . A recent classification lists four types of disease: gastric , intestinal , ectopic ( extragastrointestinal ) , and allergic ( gastro-allergic ) anisakiasis [16–18] . Gastric anisakiasis is characterized by redness and swelling at the site of larval migration , hemorrhages , erosive gastritis , and mucosal ulceration with mixed inflammatory infiltrate dominated by eosinophilic granulocytes [17] . Intestinal anisakiasis is characterized by mucosal edema and luminal stenosis; rarely intestinal obstruction and intussusception occur [17 , 19] . Chronic infection can result in formation of abscesses , ulcers , and eosinophilic granulomas forming around dying larva [17 , 20] . Such histopathological changes are caused by mechanical tissue damage during larval migration as well as secreted serine proteases and metalloproteases that degrade collagen and glycoproteins to facilitate larval migration [21 , 22] . In addition , such changes may also arise from the cytotoxic effect of eosinophilic granulocytes that accumulate in high numbers around the lesion [16] . Helminth infections in both humans and animals are characterized by induction of the T helper 2 ( TH2 ) immune response , with the exception of species of the genus Trichuris , which elicit mixed TH1/TH2 response [23] . Although different types of parasite induce the TH2 response , effects of such a response can differ greatly either resulting in worm expulsion or downregulation of TH1 and TH17 thus preventing associated immunopathology [24] . The hallmark of the TH2 response is the production of type 2 cytokines such as interleukin-4 ( IL-4 ) , IL-5 , IL-9 , and IL-13 , goblet cell ( GC ) hyperplasia and subsequent mucus production , mastocytosis , expansion of eosinophils and basophils as well as alternatively activated macrophages ( AAM ) , which contribute to fibrosis and tissue repair [25–27] . Although CD4+ cells are central players in the TH2 response , basophils and type 2 innate lymphoid cells are considered important primary sources of IL-4 and IL-13 , respectively . These cytokines act in autocrine fashion to stimulate TH2-cells to reinforce their own production [25 , 28] . The majority of parasites establish chronic infections by downregulating host immune response , which is also beneficial for the host , as it diminishes related immunopathology [23] . In contrast , zoonotic infections where humans are accidental hosts , such as with anisakiasis , can result in more pronounced immune response and associated pathology [29] . Immune responses to pathogens can be regulated by several mechanisms , either post-transcriptionally via non-coding RNAs or pre-transcriptionally by modifying DNA ( acetylation , methylation ) . MicroRNAs ( miRNAs ) are a class of small non-coding RNAs , ~22 nt long , involved in post-transcriptional regulation of gene expression . [30 , 31] . A single miRNA can regulate more than one transcript and numerous miRNAs can regulate a single transcript . The involvement of specific miRNAs in immune responses to intracellular [32] and extracellular parasites [33] has been documented . DNA methylation represses gene expression by adding methyl groups to CpG dinucleotides in promoter regions of genes; thus , preventing binding of transcription factors [34] . To date , research on murine models of anisakiasis has focused mainly on chronic infection and allergic reactions , leaving a gap in our knowledge of the early immune response to Anisakis infection . Although reports of histopathological changes induced by Anisakis spp . larvae exist both from experimental anisakiasis [35] as well as human clinical cases ( e . g . [36] ) , such studies employed only conventional histological techniques ( i . e . hematoxylin-eosin staining ) lacking immunophenotyping of cellular infiltrate or the ultrastructural details of observed histopathological changes . Therefore , the aim of this study was to characterize the early immune response to Anisakis infection in rats both at cellular and molecular levels by i ) assessing histopathological changes using three different approaches ( i . e . , classic histopathological analysis , immunofluorescence and transmission electron microscopy ) ; ii ) measuring expression of a set of common inflammatory markers and their putative regulatory miRNAs in tissues where larval migration was observed and , iii ) screening for changes in global DNA methylation in infected tissues .
In vivo experiments on rats were performed at the University of Split Animal Facility ( permit number HR-POK-19 ) . The study was approved by Veterinary and Food Safety Authority , Ministry of Agriculture of the Republic of Croatia ( permit number EP 18-2/2016 ) and Ethics Committee of the University of Split , School of Medicine ( permit number 003-08/18-03/0001 ) . All animal care and use protocols applied in this study adhere to Section three—Protection of animals used for scientific purposes of the Animal Protection Act of the Republic of Croatia ( NN 102/17 ) , which implements guidelines of the Directive 2010/63/EU of the European Parliament and of the Council on the protection of animals used for scientific purposes ( SL L 276 , 20 . 10 . 2010 ) . Animals were raised and housed in plastic cages with sawdust bedding in a controlled environment ( temperature 22 ± 1°C , 12 h dark/light cycle , food and water ad libitum ) and were deprived of food 12 h prior to experimental infection . Experimental Anisakis infection was performed on 10 male Sprague-Dawley rats ( average weight 360 . 2 g ± SD 86 . 67 g ) and tissue sampling for target gene and microRNA expression and methylated DNA quantification was set to 6 h post-infection ( p . i . ) based on the results from preliminary experiment performed on 15 female Sprague-Dawley rats ( average weight 197 g ± SD 13 . 6 g ) . The preliminary experiment has been previously described in detail in [37] . The samples for histopathological analysis , immunofluorescence , and transmission electron microscopy were collected during the preliminary experiment ( 6 , 10 , 24 , 48 , 72 h p . i . ) and used exclusively in this study . Prior to experimental infection , animals were anesthetized with a mixture of anesthetic and analgesic by intraperitoneal injection , 50–100 mg/kg Ketaminol ( Richter Pharma AG , Wels , Austria ) , and 5–10 mg/kg Xylapan ( Vetoquinol UK Ltd , Buckingham , UK ) . When no toe pinch reflex was detected , each animal was orally intubated with 10 live Anisakis larvae , collected from blue whiting Micromesistius poutassou , according to a previously described protocol [35] . Animals were euthanized by an overdose of anesthetic ( > 150 mg/kg ) followed by decapitation to confirm death . Each animal was dissected and inspected for lesions caused by migrating Anisakis larvae . Tissues with visible lesions with or without larva migrans were sampled and stored appropriately for downstream analyses ( S1 Table ) . Additionally , adjoining unaffected tissue was sampled and served as an internal control in order to reduce interindividual variability . No external control ( uninfected , “healthy” rats ) was used in this study , considering that principal component analysis ( PCA ) results from our previous study showed higher variability between such specimens within the same experimental groups [37] . For histopathological analysis , tissue samples of Anisakis lesions from preliminary experiments were fixed in 4% paraformaldehyde on ice , processed using standard techniques , paraffin-embedded and cut to 5 μm thickness . Tissue sections were HE stained , mounted in Canada balsam ( Sigma , St . Louis , MO , US ) , coverslipped and evaluated for histopathological changes using an Olympus CX40 microscope ( Olympus Corp . , Shinjuku , Tokyo , Japan ) . Photos were captured with an Olympus Camedia camera ( Olympus Corp . ) and assembled with Photoshop CS5 software ( Adobe Systems , San Jose , CA , USA ) . Cluster of differentiation 3 ( CD3 ) , inducible nitric oxide synthase ( iNOS ) and S100A8/A9 ( calgranulin A and B , calprotectin as heterodimer; migration inhibitory factor-related protein 8 and 14 ) were chosen as markers for immunofluorescent evaluation of Anisakis lesions . CD3 is a pan T-cell marker , which associates with the T-cell receptor ( TCR ) to activate different subsets of T cells [38] . iNOS is one of three nitric oxide synthases ( the other two being neuronal [nNOS] and endothelial [eNOS] ) that is not constitutively expressed but strongly upregulated in many cell types in response to various pathogens [39] . S100A8 and A9 are alarmins or DAMPs ( damage-associated molecular pattern proteins ) that are secreted from damaged cells or activated granulocytes and monocytes , and play a pivotal role in mediating the inflammatory response [40] . Immunofluorescence staining was performed on 4% paraformaldehyde fixed , paraffin-embedded tissue samples from the preliminary experiment . Tissue sections ( 5 μm ) were placed on FLEX IHC microscope slides ( Dako , Glostrup , Denmark ) , deparaffinized and rehydrated . Antigen retrieval was performed by incubating sections at 95°C in Dako Antigen Retrieval Solution , pH = 6 . 0 ( Dako ) for 20 min . Following cooling to room temperature , sections were incubated in 1% BSA in PBS ( Sigma , St . Louis , MO , US ) at room temperature to prevent non-specific binding . Due to high erythrocyte autofluorescence in the green channel , Image-iT FX signal enhancer ( Invitrogen , Carlsbad , CA , US ) was applied to slides stained with an Alexa Fluor488-conjugated secondary antibody for 30 min at room temperature and washed with Dako Wash Buffer ( TBS-Tween 20 ) ( Dako ) prior to BSA blocking . Sections were then incubated with rabbit anti-CD3 epsilon antibody ( 1:200 , ab49943 , Abcam , Cambridge , UK ) , rabbit anti-iNOS antibody ( 1:100 , ab15323 , Abcam ) and mouse anti-MRP8+MRP14 [B314 . 1 ( MAC 387 ) ] ( 1:100 , ab130234 , Abcam ) for 1 h at room temperature . Subsequently , sections were washed in Dako Wash Buffer and incubated with donkey anti-rabbit IgG H&L ( Alexa Fluor594 ) ( 2 μg/ml , ab150068 , Abcam ) and goat anti-mouse IgG H&L ( Alexa Fluor488 ) ( 1:200 , ab150117 , Abcam ) , respectively , for 1 h at room temperature . Finally , sections were washed with TBS , counterstained with DAPI , and coverslip mounted using Shandon Immu-Mount ( Thermo Scientific , Waltham , MA , US ) . As a negative control , primary antibodies were omitted ( Supporting information ) . Sections were inspected under a Zeiss Axio Imager M1 fluorescence microscope equipped with AxioCam Mrm Rev 3 ( Carl Zeiss AG , Oberkochen , Germany ) . Images were captured with AxioVision Rel . 4 . 7 software ( Carl Zeiss AG ) and assembled with Photoshop CS5 software ( Adobe Systems ) . For TEM , small pieces of Anisakis-induced lesions from the preliminary experiment were fixed in 4% paraformaldehyde on ice and post-fixed in 1% aqueous osmium tetroxide . In-block staining with 2% aqueous uranyl acetate was performed overnight , followed by dehydration in an ascending series of acetone and embedding in Durcupan resin ( Honeywell-Fluka , Morris Plains , NJ , US ) . Semi-thin sections were cut to 500 μm , stained with 1% toluidine blue and inspected under a light microscope for orientation . Ultrathin sections were cut at 0 . 07 μm , stained with uranyl acetate and lead citrate [41] , and inspected under a Jeol JEM-1400 TEM operating at 120 kV . Total RNA was extracted using TriReagent ( Ambion Inc . , Invitrogen , Carlsbad , CA , USA ) following the manufacturer’s protocol and dissolved in 20–40 μL of Mili-Q water ( Merck Millipore , Billerica , MA , US ) . The quantity and quality of extracted RNA were checked by spectrophotometry and 1% agarose gel electrophoresis , respectively . Prior to cDNA synthesis , RNA samples were treated with DNA-free DNA Removal Kit ( Invitrogen , Carlsbad , CA , USA ) to avoid amplification of genomic DNA in the downstream analysis . For target gene expression , cDNA was synthesized from 1 μg of total RNA using a High-Capacity cDNA Reverse Transcription Kit ( Applied Biosystems , Waltham , MA , US ) following the manufacturer’s protocol . Target genes were selected based on recently published transcriptomic data [37] and evidence for their regulation by miRNAs . The target genes are interleukin 1 beta ( Il1b ) , Il6 , Il18 , chemokine ( C-C motif ) ligand 3 ( Ccl3 ) , intracellular adhesion molecule 1 ( Icam1 ) , and matrix metallopeptidase 9 ( Mmp9 ) . Expression was quantified by real-time PCR using LightCycler 480 SYBR Green I Master ( Roche Diagnostics , Manheim , Germany ) with cycling conditions as previously described [42] and annealing set to 60°C . Vacuolar protein sorting-associated protein 29 ( Vps29 ) and glucose-6-phosphate isomerase ( Gpi ) were selected as reference genes based on their stability in recent transcriptome [37] and BestKeeper [43] analyses . Prior to real-time PCR , the template cDNA was diluted 1:20 with MilliQ water and each sample was run in duplicate . Primer3 ( v . 4 . 1 . 0 . ) web interface [44] was used to create specific primers ( Table 1 ) . For miRNA expression , cDNA was synthesized using miScript II RT Kit ( Qiagen , Hilden , Germany ) following the manufacturer’s protocol . miRNA expression was evaluated on miRNA PCR array Rat Immunopathology ( MIRN-104ZF ) plates with QuantiTect SYBR Green Master Mix ( Qiagen , Hilden , Germany ) . All real-time PCR experiments were run on a LightCycler 480 II System ( Roche Diagnostics , Rotkreuz , Switzerland ) . For DNA extraction , tissue samples were stored in 96% EtOH at 4°C until analysis . Total genomic DNA was extracted using a DNeasy Blood and Tissue Kit ( Qiagen , Hilden , Germany ) . The quality and quantity of extracted DNA were checked as mentioned previously . Global DNA methylation was quantified using a Methylated DNA Quantification Kit ( Colorimetric ) ( ab117128 , Abcam , Cambridge , UK ) , following the manufacturer’s protocol . The plate was read on a Microplate Photometer MPP-96 ( Biosan , Riga , Latvia ) . Log2 transformation and differential expression analysis of miRNA and target genes was performed with the limma package [45] for R ( ver . 3 . 4 . 2 ) [46] . Target genes expression was normalized on individual basis against the geometric mean of two housekeeping genes ( Vps29 and Gpi ) . Sample RN10_3K ( intestine , uninfected ) was excluded from further expression analysis since explorative analysis showed this sample to be an outlier . Expression analysis for both miRNA and target genes was run separately for each tissue using a paired design , which considers that uninfected ( control ) tissue was sampled from the same animal as infected tissue . Fold changes were calculated relative to the uninfected tissue as a group . Fold-change greater than 2 ( log2FC ≥ 1 ) was considered biologically significant , while statistical significance was set to adjusted p-value < 0 . 05 . Biological significance is regarded as a change high enough to result in certain outcome , i . e . change in gene expression that could result in enough protein secretion to manifest a certain effect . Target genes and miRNA expression analysis results were visualized with ggplot2 [47] and gplots [48] for R , respectively . To test for differences in global DNA methylation between infected and uninfected tissue of rat stomach and intestine regarding alterations in global DNA methylation , Wilcoxon signed-rank test was performed in R . This non-parametric statistical hypothesis test is suitable for comparison of mean ranks between paired samples , as is the case in our sampling design . It does not require the assumption of a Gaussian distribution .
Larval penetration caused moderate to severe submucosal hemorrhage in stomach ( Fig 1A ) , with focal to confluent fibrinoid necrosis in rare cases ( Fig 1A inset ) . Moderate to abundant mixed inflammatory infiltrate composed of neutrophils , macrophages , lymphocytes , and eosinophilic granulocytes was seen . In animals euthanized at the first two time points ( 6 h and 10 h p . i . ) , neutrophils were the predominant cell population in inflammatory infiltrate , with scant eosinophilic granulocytes interspersed between zymogenic cells and muscularis mucosae . Lesions from animals euthanized at later time points ( 24 h , 48 h and 72 h p . i . ) were characterized by more prominent mucosal and submucosal eosinophilic granulocyte infiltration ( Fig 1B and 1B inset ) with occasional neutrophils . In the intestine , Anisakis spp . larvae were found penetrating submucosal tissue and causing severe hemorrhage with perivascular fibrinoid necrosis ( Fig 1C ) . Scant inflammatory infiltrate composed of neutrophils , macrophages , and lymphocytes was seen in submucosal tissue , while the mucosal layer was infiltrated with scant eosinophilic granulocytes ( Fig 1C inset ) . In contrast to the intestine , moderate to extensive hemorrhage was seen in the caecum ( Fig 1D ) with more pronounced inflammatory infiltrate composed of neutrophils , macrophages , and lymphocytes and eosinophilic granulocytes infiltrating both mucosal and submucosal layers ( Fig 1D inset ) . Larval migration through muscle tissue resulted in moderate hemorrhage in epimysial and perimysial connective tissue ( Fig 1E ) accompanied by mixed inflammatory infiltrate dominated by neutrophils and macrophages ( Fig 1E inset ) , occasional eosinophilic granulocytes and scant lymphocytes . Muscle fibers showed signs of coagulative necrosis . Representative micrographs of uninfected control tissue are presented in S1 Fig . Immunofluorescent labeling of lesions caused by migrating Anisakis larvae showed positive staining for all three antibodies applied ( i . e . anti-CD3 , anti-iNOS , and anti-MRP8+MRP14 ( anti-S100A8/A9 ) ) . Anti-CD3 staining produced moderate to strong signal with thin ring-like cytoplasmic staining corresponding to scarce cytoplasm of lymphocytes . CD3+ cells were most abundant in the stomach , usually seen in groups within or adjacent to larger blood vessels ( Fig 2A , 2B , and 2C ) , indicating recent and ongoing extravasation . Occasionally , single CD3+ cells were observed deeper in the submucosa within large hemorrhages or very rarely in the lamina propria . In the intestine , only a few cells moderately expressing CD3 were observed in the submucosa with similar localization as in the stomach ( Fig 2D , 2E , and 2F ) ( i . e . within or adjacent to blood vessels ) . The smallest number of CD3+ cells , but with strongest expression , was found in muscle tissue . Single cells were occasionally seen within blood vessels or infiltrating perimysial connective tissue ( Fig 2G , 2H , and 2I ) . Uninfected control tissue and negative stain control tissue are presented in S2 Fig and S3 Fig , respectively . Anti-iNOS immunolabeling produced moderate to intense signal , showing either granular or diffuse cytoplasmic staining . In stomach , individual or small groups of iNOS+ cells could be seen throughout extensive submucosal hemorrhages ( Fig 3A , 3B , and 3C ) . Most of these cells were characterized by moderate granular cytoplasmic and perinuclear staining , with occasional cells yielding stronger signals . Based on cellular and nuclear morphology the most abundant iNOS+ cells appeared to be neutrophils and fibrocytes , however , no staining with specific markers has been performed to confirm cell types . In intestine , similarly to the stomach , iNOS+ cells were observed throughout the submucosal hemorrhage but tended to group around the migrating Anisakis larva . Again , various cell types were iNOS+ but yielded a stronger signal than their counterparts in stomach ( Fig 3D , 3E , and 3F ) . The highest number of iNOS+ cells , showing the strongest expression , was seen in muscles . Most of the cells showed strong granular or diffuse cytoplasmic staining ( Fig 3G , 3H , and 3I ) . As in the stomach and intestine , the most abundant iNOS+ cells appeared to be neutrophils and fibrocytes . A low number of cells , most likely mononuclear phagocytes , showing very strong , diffuse cytoplasmic staining could be seen randomly dispersed in the perimysium or adjacent to blood vessels . Uninfected control tissue and negative stain control tissue are presented in S4 Fig and S5 Fig , respectively . Anti-MRP8+MRP14 ( anti-S100A8/A9 ) staining yielded moderate to very strong granular or diffuse cytoplasmic signals . In the stomach of animals euthanized at earlier time points , a fairly low number of cells with moderate to strong expression was seen in muscularis externa ( Fig 4A , 4B , and 4C ) together with , most likely , rare fibrocytes in the submucosa and eosinophils in the lamina propria , respectively; however no staining with specific markers has been performed to confirm cell types . In animals euthanized at later time points , groups of eosinophils with high expression could be seen in the lamina propria . In the intestine , a low number of MRP8+MRP14+ cells were seen , mostly fibrocytes and endothelial cells ( Fig 4D , 4E , and 4F ) . The highest number of MRP8+MRP14+ cells with strongest expression was seen in muscle . Numerous immunoreactive cells , including most likely fibrocytes , eosinophils , and neutrophils , yielding strong ( rarely very strong ) signals , could be seen in perimysium and epimysium . However , considerably stronger expression , occasionally exceptionally strong , was seen in myocytes , which were also the most abundant MRP8+MRP14+ cells ( Fig 4G , 4H , and 4I ) . In longitudinal muscle fiber section , strong granular sarcoplasmic staining could be seen , while in cross-sections , cells yielded strong diffuse sarcoplasmic signals . Uninfected control tissue and negative stain control tissue are presented in S6 Fig and S7 Fig , respectively . TEM analysis of Anisakis-induced lesions revealed ultrastructural details of inflammatory infiltrate and tissues changes . As seen in paraffin sections , gastric lesions were characterized by hemorrhage and abundant neutrophil and macrophage infiltration ( Fig 5A and 5B ) . The characteristic “coffee bean” structure of cytoplasmic granules evidenced the presence of eosinophils in inflammatory infiltrate and a lack of mast cells ( Fig 5B ) . Macrophages with engulfed erythrocytes could occasionally be seen . In the intestine , neutrophils and eosinophils were seen mixed with erythrocytes and interspersed between bundles of collagen fibers . A single plasma cell , with elaborate endoplasmic reticulum , was seen associated with several eosinophils ( Fig 5C ) . However , an unexpected finding of numerous fungus-like organisms ( Fig 5C inset ) with intraluminal and submucosal localization was noticed , some of which were endocytosed by eosinophils . The organisms were elongated and round in cross-section , several micrometers long ( up to 9 μm ) and with electron light intracytoplasmic granules . Similarly , neutrophil , macrophage , and eosinophil infiltration ( lacking mast cells ) was observed in caecum ( Fig 5D ) . In muscle , several areas of structurally fragmented and necrotic fibers were observed ( Fig 5E ) . These areas were heavily infiltrated with a mixture of neutrophils , macrophages , and eosinophils ( Fig 5E ) , some with expelled granules . The expression of target genes was examined in stomach , intestine , and visceral adipose tissue . Out of six tested target genes , five were differentially expressed in stomach , none in intestine , and all six in visceral adipose tissue . In stomach , five genes were upregulated and only Mmp9 was negligibly downregulated ( Fig 6A ) . The highest fold change was observed for Il6 , followed by Ccl3 , Il1b , Icam1 , and Il18 ( Fig 6B ) . Biologically significant induction was recorded for Il6 , Ccl3 , Il1b , and Icam1 . In intestine , four genes were upregulated , with Il18 being weekly downregulated , and Icam1 showing no change in expression ( Fig 6A ) . The highest fold change was observed for Il6 , which was the only biologically significant induction among the target genes ( Fig 6B ) . In visceral adipose tissue , all target genes were upregulated ( Fig 6A ) with highest fold-change observed for Ccl3 , followed by Il1b , Il6 , Mmp9 , Icam1 , and Il18 ( Fig 6B ) . Biologically significant induction was recorded for all genes except for Il18 . miRNA expression was examined in stomach and intestinal tissues ( Fig 7A ) . In stomach , two miRNAs were differentially expressed ( rno-miR-451-5p and -223-3p ) , both displaying statistical and biological significance ( Fig 7A and 7B ) . Similarly , two miRNAs were differentially expressed in intestine ( rno-miR-451-5p and -672-5p ) ( Fig 7A and 7B ) , also displaying statistical and biological significance . When the non-adjusted p-value was considered , four additional miRNAs were differentially expressed in stomach; three were upregulated ( rno-miR-142-3p , -142-5p , and -18a-5p ) and one was downregulated ( rno-miR-205 ) ( Fig 7B ) . In intestine , three additional miRNAs were differentially expressed when the non-adjusted p-value was considered; two were upregulated ( rno-miR-363-3p and -196a-5p ) and one was downregulated ( rno-miR-298-5p ) ( Fig 7B ) . Rno-miR-451-5p was the only differentially expressed miRNA in both stomach and intestine . Total methylated DNA was quantified in stomach and intestine . No statistically significant difference ( at p < 0 . 05 ) in DNA methylation was detected between infected ( mean 2 . 56 ng , SD ± 1 . 03 ng ) and uninfected ( mean 3 . 08 ng , SD ± 1 . 39 ng ) stomach tissues or infected ( mean 4 . 25 ng , SD ± 1 . 80 ng ) and uninfected ( mean 4 . 90 ng , SD ± 2 . 25 ng ) intestine tissue of rats , respectively .
Here we present the results of a comprehensive cellular and molecular analysis of the early rat immune response to experimental Anisakis spp . infection . A limited number of studies have tackled the pathology in this host-parasite system , concentrating on general histopathological analyses ( HE staining , trichrome staining ) [35] and more recently tissue transcriptomics [37] and miRNA expression analysis in sera of infected animals [49] . To the best of our knowledge , this is the first comprehensive histopathological analysis of Anisakis lesions inferred by classic histopathology , immunofluorescence ( IF ) , and transmission electron microscopy ( TEM ) . In addition , for better insight into the underlying molecular mechanisms of the visible changes , we validated the histology findings by quantification of target genes and miRNA expression in tissues affected by larval migration , and the screening of global DNA methylation levels in Anisakis-infected vs . uninfected tissues . Histopathological analysis of Anisakis lesions revealed an acute and generally strong tissue inflammatory response , mostly dominated by neutrophils and macrophages . Both stomach and intestine tissues were hallmarked by extensive submucosal hemorrhages as well as vascular wall and perivascular necrosis , which facilitated leukocyte extravasation through increased vascular permeability . In addition to larva-induced trauma of the tissues accounting for the hemorrhages , Anisakis spp . larvae also secrete various proteolytic enzymes [22 , 50] that degrade surrounding host tissue enhancing larval movement and leading to cellular necrosis . Necrosis is a form of uncontrolled cell death that leads to leakage of various intracellular components , which can further stimulate the inflammatory response [51] . However , the abundance of the observed inflammatory infiltrate differed between stomach and intestine , being markedly elevated in the stomach . The rationale for a such discrepancy could lie in earlier larval penetration through the stomach compared to the intestinal wall , allowing more time for immune cell infiltration of the affected tissues . A similar predominance of neutrophil and macrophage lineages was observed in affected muscle tissue , but with comparatively less inflammatory infiltrate than in stomach . In addition to the later occurrence of larval migration through muscle tissue compared to stomach and consequent later mobilization of immune cells , it appears that the intact blood vessels in the muscle being less permeable to leukocyte extravasation could have added to the infiltrate scarcity . Although helminth infections are usually associated with a marked increase in eosinophil count ( eosinophilia ) [52] , we observed increased numbers of infiltrating eosinophils only in tissues from animals euthanized at later time points ( 24h , 48 h , and 72h p . i . ) . Pronounced eosinophil infiltration and subsequent formation of eosinophilic granuloma have been recorded in many cases of human anisakiasis [20 , 36] . However , eosinophilia is usually not constant at the beginning of clinical manifestations in humans [53] suggesting that the small number of observed eosinophils in rat could be simply due to tissue sampling before the onset of eosinophil infiltration . Nevertheless , eosinophil degranulation occurred early in the course of infection , as revealed by TEM , suggesting that the initial priming has been achieved Neutrophils are the first cellular component to respond to inflammation and tissue damage , reacting to a wide range of PAMPs or DAMPs , including large and antigenically complex metazoan pathogens . Neutrophil-mediated recruitment of other granulocytes and monocytes at the infection site [54 , 55] supports our findings . Macrophages have pleiotropic functions , both as immune effector cells and in maintaining tissue homeostasis , including phagocytosis of erythrocytes and clearing of cellular debris [56] . During several helminth infections , a subset of macrophages termed alternatively activated macrophages ( AAMs ) has been observed [26] . These macrophages are activated in response to TH2 cytokines , IL4 and -13 , and are characterized by the production of several enzymes such as arginase-1 , which is involved in collagen deposition [26 , 56] . Although their function as effector cells in worm expulsion has been demonstrated in mice infected with the natural gastrointestinal helminth Heligmosmoides polygyrus [57] , AAM function in wound healing and tissue remodeling caused by large metazoan parasites could be more important than host defense per se [56] . Indeed , in addition to neutrophils , we observed the presence of numerous macrophages . Both cell lineages were associated with necrotic tissue in stomach and muscle . Likewise , TEM revealed the presence of numerous macrophages with engulfed erythrocytes . This high number of macrophages and neutrophils , especially , could also account for the later onset of eosinophil tissue infiltration in the early phase of infection . In the case of Strongyloides stercolaris , another intestinal tissue-dwelling nematode , macrophages and neutrophils alone can kill its larvae if soluble factors from either cell type are present [58] . A similar succession of large numbers of neutrophils gradually being replaced by macrophages and eosinophils has also been seen in early Ascaris spp . infection in mice [59] , although the comparison with Anisakis needs to be taken cautiously , as the early infection phase of the former occurs in lungs . Immunofluorescence of rat sections clearly demonstrated early ( 6 h p . i . ) infiltration of CD3+ cells in Anisakis lesions . CD3 is a pan T-cell marker , which functions as a TCR co-receptor during activation of different T-cell subsets . A lower number of CD3+ cells was seen in both intestine and muscle compared to stomach , which is consistent with the later occurrence of larval migration through these tissues . Interestingly , compared to a single CD3+ cell in muscle , the highest number of these cells was seen in the intestine , probably due to the lesion’s proximity to one of the intestinal lymphoid follicles . Helminth infections are typically characterized by strong polarization of TH2 immune response [23] . However , in Trichuris muris infections , the worm burden can influence the polarization towards either TH1 or TH2 , such that a reduced worm burden promotes the TH1 response [60] . Furthermore , in patients with trichinellosis , about half of the peripheral blood mononuclear cells ( PBMC ) display a TH2 profile with the other half displaying either TH0 or TH1 profiles [61] . In the blood of Anisakis simplex-sensitized patients , both TH1 and TH2 cytokines have been recorded . However , a TH1-dominant response is characteristic for patients with gastrointestinal ( GI ) symptoms [62] . Whether such peripheral immune responses reflect the localized response in the GI tract still remains to be elucidated , requiring detailed immunophenotyping of the infiltrating T-cell populations in Anisakis lesions . Immunofluorescent iNOS labeling showed the presence of different iNOS+ cells in all three tissues examined . iNOS is expressed by a number of cells such as neutrophils , macrophages , and dendritic cells . High levels of nitric oxide ( NO ) are produced by macrophages in response to microbial products and TH1 cytokines [39] , as well as protozoan [63] and metazoan parasites [64 , 65] . In addition to antimicrobial and antiparasitic effects , NO has cytotoxic effects [39 , 66] , which could partially account for the observed Anisakis changes . Interestingly , the strongest iNOS expression was seen in affected muscle in several cell types including neutrophils , fibrocytes , and phagocytes both in epimysium near the site of larval penetration and distantly in the perimysium . High macrophage iNOS production was also recorded in Trichinella spiralis-infected mice , adjacent to encysted larva and surrounding muscle fibres , being most pronounced in the late phase of muscle infection [67] . Being able to cross-infect rodents , Trichinella spp . naturally reside within their striated muscles , in contrast to Anisakis for which rats and humans are naïve accidental hosts . This suggests that , being in an unfavorable environment such as striated muscle , Anisakis spp . could induce an early explosive iNOS expression like Trichinella , however , with substantial temporal difference . S100A8 and S100A9 are alarmins or DAMPs , which are constitutively expressed in neutrophils , monocytes , and dendritic cells but can also be secreted from activated or damaged macrophages , endothelial cells , fibroblasts , and keratinocytes [40 , 68] . We observed the presence of S100A8/S100A9+ cells in all three tissues examined , with the highest number of cells and the strongest expression in muscle . Unfortunately , with the applied antibody , we were not able to discern between the two proteins expressed , or their complex whatsoever . In vivo studies have demonstrated chemotactic activity of S100A8/A9 proteins as early as three hours post-stimulation [69] . The complex influences endothelial integrity through inhibition of several cell-junction proteins [70] thus facilitating leukocyte migration . Given the abundant neutrophil infiltrate and several perivascular necrotic foci in stomach , we unexpectedly identified a negligible number of cells with cytoplasmic expression of these proteins in stomach , mostly localized in the muscularis externa . This suggests that the tissue sampling ( 10 h p . i . ) coincided with the cells having already secreted the complex into the extracellular space , affecting its intracellular detection . The detection was further hindered by high autofluorescence of the hemorrhages . Similarly , a relatively small number of S100A8/A9+ cells were detected in the intestine , mostly expressed by endothelial cells and possibly occasional fibrocytes . Whether the low number of positive cells is due to early sampling ( relative to larval migration time ) remains unclear , considering the overall moderate inflammatory infiltrate . In muscle tissue , we found various S100A8/A9+ cells , most likely fibrocytes , neutrophils , and occasional eosinophils , which is consistent with previous reports [68 , 71] . However , the very strong expression in myocytes is puzzling . While expression of these proteins can be induced in murine vascular smooth muscle cells or cardiomyocytes following inflammatory stimuli such as bacterial infection [72] , we failed to find references demonstrating expression of these markers at the protein level in skeletal muscle . However , in a recently published transcriptome analysis of A . pegreffii-infected rats , S100A8 and S100A9 were among the most highly expressed genes [37] . S100A8/A9 protein expression during helminth infection ( particularly in neutrophils ) , has been demonstrated in lung inflammation in mice following infection with Litomosoides sigmodontis L3 larvae [73] as well as at the mRNA level in sheep infected with the digenean Fasciola hepatica [74] . Acute inflammation , such as in a rat model of Anisakis spp . infection , is characterized by production of numerous cytokines and signaling molecules that govern the immune response and wound healing after tissue damage . Here , we assessed the expression of several common inflammatory markers ( i . e . , Il1b , Il6 , Il18 , Ccl3 , and Icam1 , as well as Mmp9 , which is involved in tissue remodeling [75] ) . Interestingly , while all markers except Mmp9 were differentially expressed in adipose and stomach tissues , respectively , no difference was detected in the intestine . As mentioned earlier , severe submucosal hemorrhages were present both in stomach and intestine following larval migration . Hemorrhages have been shown to potentiate the expression and secretion of proinflammatory cytokines , including IL-1 beta and IL-6 , as well as delay wound healing following trauma [76] . In contrast , the lack of significant induction of these genes in the intestine is likely due to the early sampling in respect to larval penetration , corresponding to observed mild to moderate inflammatory infiltrate in paraffin sections . However , the possibility that proinflammatory cytokines other than those evaluated are induced earlier in response to Anisakis migrating larvae should not be excluded . Surprisingly , all six genes were differentially expressed in visceral adipose tissue , with all but Il18 displaying a biologically significant induction . In addition to the most abundant adipocytes , adipose accumulations encompass various tissue types , endothelial cells , smooth muscle cells , and fibroblast as well as almost all immune cells including resident macrophages , neutrophils , T cells , and B cells , to name a few [77] . In addition to energy storage , adipose tissues have an important role in regulating both local and systemic inflammation through secretion of anti- and proinflammatory cytokines , including IL-1 beta , IL-6 , and IL-18 [78] , as well as MMP9 [79] . However , at this point , it remains unclear whether such strong induction of proinflammatory genes in adipose tissue is due to the mere presence of Anisakis larva , tissue damage and inflammation in adjoining tissue or both . In stomach , Mmp9 was the only gene that showed virtually no difference in expression compared to control . MMP9 is known to be involved in tissue remodeling and wound repair; however , wound healing might be delayed by hemorrhage through increased production of proinflammatory cytokines [76] . The lack of change in Mmp9 expression could be the result of delayed induction due to severe submucosal hemorrhages . In addition , MMP9 is suggested to be important for neutrophil migration through basement membrane [80] , congruent with our histology data , as we found no intraepithelial infiltration of neutrophils , likely due to no Mmp9 expression at this site . All three proinflammatory cytokines ( i . e . Il1b , Il6 , and Il18 ) were differentially expressed in stomach , although Il18 did not display biologically significant induction . Both IL-1β and IL-18 are activated and secreted from multiprotein complexes called inflammasomes , containing one of the NOD-like receptors ( NLRs ) , such as NLRP3 [81] . The difference in expression of these two genes could , therefore , be due to factors other than the mechanism of activation shown in the mouse model of Trichuris muris infection , where both the worm and its secreted products induced NLRP3-dependent production of IL-18 [82] . IL-18 induces TH1 polarization through secretion of interferon-gamma ( IFN-γ ) , except in the absence of IL-12 or IL15 , when it promotes TH2 polarization [83] . Indeed , T . spiralis-infected IL-18 knockout mice had significantly higher production of TH2 cytokines with lower worm burden , compared to wild type-mice [84] . Nevertheless , induction of Il18 and its role in driving the TH1 response could indicate that early rat responses to Anisakis spp . are of mixed TH1/TH2 type possibly with TH2 predominance , similar to the intestinal phase of mouse T . spiralis infection [85] . IL-1β is a major endogenous pyrogen , which activates neutrophils and macrophages for phagocytosis of invading pathogens and production of oxygen and nitrogen radicals . In addition , it stimulates production of other proinflammatory cytokines , such as tumor necrosis factor α ( TNFα ) and IL-6 [81] . However , in the context of helminth resistance , IL-1 beta can have contrasting effects . While it attenuates the protective TH2 response and promotes parasite chronicity in H . polygyrus infection [86] , it induces the TH2 response and parasite expulsion in T . muris infection [87] . How IL-1 beta affects immunity during Anisakis infection remains unclear because the parasite fails to achieve adult stage in the accidental host , which would warrant chronic inflammation , or has a low chance of re-infecting the same accidental host over a longer time span . Remarkably strong expression was observed for Il6 ( logFC = 4 . 07 ) . IL-6 is another major endogenous pyrogen , which plays a pivotal role in the immune response by switching from neutrophil to monocyte recruitment , skewing monocyte differentiation into macrophages and T-cell expression and differentiation to TH2 and TH17 phenotypes [88] . However , mice infected with H . polygyrus had a suppressed TH2 response and higher number of FoxP3+ cells compared to IL-6-deficient mice [89] , indicating a pleiotropic role for this cytokine in regulating the immune system . Interestingly , we noted a very strong expression of Ccl3 , which exceeded the expression of Il1b . CCL3 is a chemokine with pyrogenic properties , but more importantly is a chemoattractant for neutrophils and an activator of their oxidative burst [90] . In contrast to S100A8/A9+ cells , which were observed in small numbers , the strong Ccl3 expression likely accounted for the abundant neutrophil infiltration observed in Anisakis lesions . Although significant , Icam1 displayed moderate expression compared to the aforementioned markers . ICAM1 is expressed by endothelial cells and upon inflammatory stimulus leads to increased leukocyte transmigration and increased vascular permeability through downregulation of cell-junction proteins [91] . ICAM1 regulates adhesion of neutrophils to endothelium and their transendothelial migration via either paracellular or transcellular routes [92] . Therefore , the abundant neutrophil infiltration is likely a result of the combined effects of CCL3 and ICAM1 , although the latter showed four-fold lower expression than the former . This discrepancy in expression is possibly due to vascular wall necrosis leading to a partial loss of Icam1 mRNA . Both live L3 Anisakis larvae and protein crude extract ( CE ) , which simulates dying larva , were recently found to stimulate human monocyte-derived dendritic cells ( DCs ) to secrete CCL3 , IL-6 , soluble ICAM1 ( sICAM1 ) , and IL-1 alpha , another isoform of IL-1 with properties overlapping IL-1 beta [93] , thus , confirming our findings . MicroRNA expression analysis was performed in stomach and intestine tissues affected by Anisakis larval migration . In total , three miRNAs were differentially expressed , two in stomach ( rno-miRNA-451-5p and -223-3p ) and two in intestine ( rno-miRNA-451-5p and -672-5p ) , with rno-miRNA-451-5p being expressed in both tissues . To the best of our knowledge , this is the first miRNA expression analysis performed in tissue samples directly affected by larval migration , rather than in sera of infected animals as reported by Corcuera et al . ( 2018 ) . The authors found two differentially expressed miRNAs in rat sera , in addition to four others based on RQ values , which were different to those we identified in tissues . However , this could be attributed to the mode of infection ( live L3 vs . crude extract inoculated into gastric mucosa ) , duration of infection ( 6 h p . i . vs . 2 . 5 months ) and sample type ( affected tissues vs . serum ) . In Anisakis stomach lesions , miRNA-451 displayed particularly strong expression ( logFC = 4 . 51 , adj . p = 1 . 88e-8 ) and roughly four-fold lower expression in intestine ( logFC = 2 . 42 , adj . p = 0 . 036 ) . miRNA-451 has been shown to downregulate CD4+ T-cell proliferation , as evidenced by mice infected with Plasmodium yoelli having significantly lower numbers of CD4+ cells in addition to higher parasite burden compared to miR-451-/- mice [94] . Furthermore , miR-451 impairs neutrophil chemotaxis [95] and negatively affects secretion of IL-6 and CCL3 [96] . In stomach , we observed abundant neutrophil infiltration as well as strong expression of both Il6 and Ccl3 . IL-6 has been shown to positively regulate expression of miR-451 , potentially forming a regulatory loop that culminates in decreased IL-6 expression [96] . Therefore , the expression of miR-451 in stomach is likely the consequence of strong positive feedback of Il6 expression , aimed at the downregulation of over expressed Il6 . Interestingly , mild inflammatory infiltrate coupled with the lack of statistically significant expression of the target genes was seen in intestine , although miR-451 was significantly expressed . Nevertheless , we observed biologically significant induction of Il6 ( logFC = 1 . 47 ) , which might have been sufficient to drive initial expression of this particular miRNA . Although being expressed eight-fold lower than miR-451 , miR-223 showed significantly higher expression in stomach compared to control . It was reported that absence of miR-223 drives accelerated recruitment of myeloid cells ( i . e . polymorphonuclear leukocytes and mononuclear phagocytes ) at the site of infection , and subsequent tissue damage [97] . Therefore , we hypothesize that induction of this miRNA serves to limit neutrophil infiltration and consequent tissue damage , and gradual substitution with other immune cells . Furthermore , miRNA-223 has also been reported to directly target Ccl3 and Il6 [97] . In addition , miRNA-223 has been reported to negatively regulate the production of IL-1 beta by repression of the NLRP3 inflammasome [98] . Taken together , it appears that miRNA-223 could play an important role in shaping Anisakis pathology despite substantially lower expression compared to miR-451 . However , IL-6 itself is able to downregulate expression of miRNA-223 [99] . Therefore , the difference in expression between miRNA-223 and miR-451 could be attributed to strong Il6 expression . Intestinal expression of miR-672 during Anisakis infection remains enigmatic , as does its role in regulation of mRNA in general . Namely , data on the function of miR-672 is scarce , except for evidence of its modulation in a murine model of acute and chronic asthma [100] . Considering the substantial difference in inflammatory infiltrate between stomach and intestine and modest downregulation of this miRNA in stomach , we can only assume that the quantity and type of immune cells infiltrating the two organs account for miR-672 induction in the intestine . In addition to post-transcriptional regulation via miRNAs , gene expression can be regulated by chemical modifications , such as DNA methylation , which prevents binding of transcription factors by introducing changes in promoter regions [34] . Moreover , aberrant DNA methylation has been associated with several types of cancer due to repression of tumor-suppressor genes [101] . In this study , we screened for changes in global DNA methylation status and found no significant differences between infected and uninfected tissue . We argue that despite our results , DNA methylation could indeed have a role in the immune response to Anisakis infection , but the effect is probably reserved for prolonged Anisakis infection or to specific genes non-discernible by performed design . In a murine model of urogenital schistosomiasis , changes in DNA methylation levels coincided with urothelial hyperplasia , a key preneoplastic lesion leading to schistosomal bladder cancer [102] . Unlike S . haematobium , Anisakis spp . are not considered carcinogenic , but their potential role in carcinogenesis has been controversial ever since it was suggested that A . simplex might be a co-factor for the development of gastric cancer [103] . Interestingly , Sonoda et al . [104] summarized 27 cases reporting attachment of Anisakis larva to GI cancer , and given the role of methylation in carcinogenesis , this interaction warrants further research . Taken together , our results show that Anisakis spp . trigger a rapid and robust immune response in rats , characterized by abundant neutrophil and macrophage infiltration and significant induction of Il6 and Il1b proinflammatory cytokines and the chemokine Ccl3 , which were among the most highly expressed genes of more than 1300 differentially expressed genes in the transcriptome of Anisakis pegreffii-infected rats [37] . Moreover , three miRNAs were differentially expressed , indicating their involvement in particular mRNA expression profiles measured in affected tissues . However , the functional role of these miRNAs in Anisakis infection remains unknown . In contrast , no changes in DNA methylation status were observed between infected and uninfected tissue tissues . Establishing chronic infection and employing different methodological approaches to study the functional role of miRNAs and DNA methylation will help to address some of the controversies surrounding Anisakis infection in humans . | Anisakiasis is a zoonotic disease ( infection transmitted between animals and humans ) contracted by consumption of raw or undercooked seafood contaminated with Anisakis spp . nematode larvae . Anisakiasis usually presents with variable and unspecific gastrointestinal and/or allergic symptoms , which accounts for the high rate of misdiagnosed cases . Due to changes in dietary habits , such as eating raw or undercooked seafood , anisakiasis is considered an emerging public health problem . Despite the increase in number of reported cases worldwide , mechanisms of immune response to this unspecific human pathogen are poorly known . We have shown that in experimentally infected rats , Anisakis larvae cause severe hemorrhages and necrotic changes of affected tissues in the early phase of infections . Neutrophils and macrophages were abundantly present in tissue lesions , while eosinophils , hallmark of helminth infections , were scarcely present . We have also demonstrated particularly strong expression of several inflammatory genes . Moreover , we give for the first-time insight into putative regulatory mechanism mediated via a distinct class of RNA molecules . Our study may provide new opportunities for better understanding of cellular and molecular response to Anisakis spp . , aiming at development of more specific therapeutics and alleviation of pathologies associated with Anisakis spp . infection . | [
"Abstract",
"Introduction",
"Materials",
"and",
"methods",
"Results",
"Discussion"
] | [
"blood",
"cells",
"invertebrates",
"medicine",
"and",
"health",
"sciences",
"immune",
"cells",
"pathology",
"and",
"laboratory",
"medicine",
"natural",
"antisense",
"transcripts",
"gene",
"regulation",
"immunology",
"animals",
"developmental",
"biology",
"micrornas",
"signs",
"and",
"symptoms",
"stomach",
"anisakis",
"neutrophils",
"digestive",
"system",
"white",
"blood",
"cells",
"inflammation",
"animal",
"cells",
"gene",
"expression",
"life",
"cycles",
"immune",
"response",
"gastrointestinal",
"tract",
"biochemistry",
"rna",
"eukaryota",
"diagnostic",
"medicine",
"anatomy",
"cell",
"biology",
"nucleic",
"acids",
"genetics",
"nematoda",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"macrophages",
"non-coding",
"rna",
"larvae",
"organisms"
] | 2019 | Interplay between proinflammatory cytokines, miRNA, and tissue lesions in Anisakis-infected Sprague-Dawley rats |
Understanding how dynamical responses of biological networks are constrained by underlying network topology is one of the fundamental goals of systems biology . Here we employ monotone systems theory to formulate a theorem stating necessary conditions for non-monotonic time-response of a biochemical network to a monotonic stimulus . We apply this theorem to analyze the non-monotonic dynamics of the σB-regulated glyoxylate shunt gene expression in Mycobacterium tuberculosis cells exposed to hypoxia . We first demonstrate that the known network structure is inconsistent with observed dynamics . To resolve this inconsistency we employ the formulated theorem , modeling simulations and optimization along with follow-up dynamic experimental measurements . We show a requirement for post-translational modulation of σB activity in order to reconcile the network dynamics with its topology . The results of this analysis make testable experimental predictions and demonstrate wider applicability of the developed methodology to a wide class of biological systems .
Uncovering how regulatory networks shape the dynamical properties of cellular responses to external stimuli is one of the ultimate goals of system biology . Despite our successes in mapping and modeling gene regulatory networks for a wide variety of model systems , only a handful of generalizable rules relating network architecture to its dynamic performance have been formulated . These rules are often called evolutionary design principles of biochemical networks [1] . Developing new approaches to find these design principles should allow us to extend our understanding of biological network dynamics from a few case studies to a wide variety of model systems . Here we formulate one such approach and apply it to a network controlling the response of Mycobacterium tuberculosis , the causative agent of tuberculosis ( TB ) , to hypoxic stress . With almost one-third of the world population infected , TB remains a major public health threat [2] . M . tuberculosis survives stress conditions induced by host immunity by undergoing major metabolic and physiological remodeling that leads to mycobacterial dormancy [3–6] . Understanding this adaptive response of the tubercle bacillus is central to our long-term ability to control the pathogen . Transcriptional networks downstream of the alternative sigma factor σE , are critical for this adaptive response . They are activated when bacteria infect host macrophages , and induce the production of virulence factors and host inflammatory responses [7 , 8] . Deletion of sigE leads to the strongest attenuation of M . tuberculosis murine infection among all accessory sigma factor mutant strains [7] . Induction of σE can be studied in vitro by exposing M . tuberculosis cells to a wide range of stressors such as hypoxia and surface or oxidative stress [7 , 8] . Due to its importance for cellular survival , σE is subjected to rather complex regulatory mechanisms at both the transcriptional and post-translational levels . Transcription of sigE is controlled by three different promoters ( P1-3 ) . P1 is responsible for basal expression of sigE under normal physiological conditions . P2 is activated in the presence of MprA , part of the MprAB two component system that can sense surface stress [9 , 10] . Interestingly , σE also activates the transcription of MprAB , forming a positive feedback loop . The last promoter ( P3 ) is transcribed by the σH-RNAP holoenzyme in response to oxidative stress . σE post-translational regulation is primarily controlled by its anti-sigma factor , RseA . RseA binds to and sequesters σE , preventing the formation of the active σE-RNAP holoenzyme . However , under stressful conditions , PknB ( a eukaryotic-like Ser/Thr protein kinase ) will phosphorylate RseA , tagging it for degradation by the ClpC1P2 proteases [10 , 11] . Recently , the transcriptional dynamics of σE and several of its regulon members following hypoxic stress have been quantified , and the networks controlling production of two critical central metabolism genes , icl1 ( Rv0467 , glyoxylate shunt ) and gltA1 ( Rv1131 , methylcitrate cycle ) , have been characterized [12] . These genes are implicated in the growth-phase-dependent metabolic adaptation of M . tuberculosis [13] and are essential for growth and persistence of tubercle bacilli in infection models [14–16] . The glyoxylate shunt is especially important because it allows M . tuberculosis to efficiently metabolize fatty acids; indeed , it has been suggested that fatty acids may be the major source of carbon and energy for tubercle bacilli in chronically infected lung tissue [16] . For icl1 ( Fig 1A ) , transcription requires both σB , an alternative sigma factor transcribed under σE control , and a σB-regulated transcription factor , named lrpI ( Rv0465c , local regulatory protein of icl1 ) [12] . This network motif–coherent feedforward loop–is a common motif of bacterial regulatory networks that is known to produce delays in responses to step-up inputs , to filter transient stimuli , and to increase network sensitivity [1 , 17] . Notably , in the case of icl1 , the resulting activation dynamics is different from that is seen for other coherent feedforward loops . Following gradual depletion of oxygen over the course of 3 days , which leads to σE activation , icl1 is transiently induced on day 4 , and then decreases to pre-induction levels by day 5 ( Fig 1B ) . A similar transient surge in icl1 has also been observed in vivo [12] . In this paper , we develop a methodology to uncover the mechanism of non-monotonic response following the monotonic dynamics of a stimulus , and apply it to the transient surge in icl1 dynamics . We first formulate a general theorem based on monotone systems theory that gives a necessary condition for this observation . Then we employ a combination of modeling , parameter optimization , and experimental tests to uncover missing interactions in the icl1 regulatory network and to make new experimentally testable predictions .
A major challenge in the quantitative analysis of biological systems is their tendency to be highly complex and non-linear , complicating analysis of system behavior . Therefore , there is a substantial need for methods that can make it possible to constrain potential network topologies based on the observed dynamics without knowledge of the underlying parameters . Here we develop a generalized theorem that gives a necessary condition for non-monotonic system dynamics , given a monotonic input signal . In this section , we semi-intuitively describe the concept and main results . A comprehensive formulation and the proof of the theorem are found in the S1 Text . A function is defined as monotonic if it is increasing or decreasing over a given domain . That is , for a monotonically increasing function , for all t1 ≥ t0 , f ( t1 ) ≥ f ( t0 ) . For monotonically decreasing functions , the sign of the inequality is flipped , i . e . f ( t1 ) ≤ f ( t0 ) . If the function is defined by a dynamical system , in the sense that it is a component of a solution of a set of ordinary differential equations , we set to identify general properties of the dynamical system that ensure monotonic increase/decrease of its components over time . In the context of biological networks , the components usually represent concentrations of biological species ( e . g . mRNA , proteins , metabolites , etc . ) . The dynamical system consists of biochemical kinetics equations for these species: ddtX1 ( t ) =F1 ( X1 ( t ) , X2 ( t ) , … , Xn ( t ) ;u ( t ) ) ( 1 ) ddtX2 ( t ) =F2 ( X1 ( t ) , X2 ( t ) , … , Xn ( t ) ;u ( t ) ) ⋮ddtXn ( t ) =Fn ( X1 ( t ) , X2 ( t ) , … , Xn ( t ) ;u ( t ) ) Here Xi ( t ) are time-dependent concentrations of relevant chemical species and Fi is the net flux into a given concentration pool , i . e . net sum of all the rates of reactions producing specie i minus all the rates of reactions that consume it . The function u ( t ) is a known time-dependent input into this biochemical system , such as an externally supplied chemical ligand or stressor . Usually , it will only directly affect one or a few network nodes . For generality we allow multiple nodes to be directly affected by u ( t ) ; these would be called the 'input variables’ . On the other hand , the last node Xn ( t ) is arbitrarily designated as the 'output variable’ . For brevity , we can represent this dynamical system in the vector-notation as ddtX ( t ) =F ( X ( t ) ;u ( t ) ) ( 2 ) where X and F are vectors with components Xi and Fi respectively . We may represent this dynamical system as a graph consisting of ( n+1 ) nodes and directed edges connecting some of the nodes ( such as the network diagram on Fig 1A ) . The nodes correspond to the input function u ( t ) and all chemical species Xi ( t ) ( i = 1…n ) . The edges correspond to biochemical interactions: the input node corresponding to u ( t ) is connected to all input variables , i . e . all variables for which ∂Fi/∂u ≠ 0 . Analogously , the node corresponding to concentration Xk will be connected by a directed edge from Xl if ∂Fk/∂Xl ≠ 0 . If the increase in the concentration of species l increases the production flux for concentration of k ( i . e . , ∂Fk/∂Xl ≥ 0 ) , then we draw an edge with an arrow ( ↓ ) or assigned parity +1 . Biologically , this corresponds to “activation” . Conversely , if j inhibits/represses species i ( i . e . , ∂Fi/∂Xj ≤ 0 ) , then we will draw an edge with blunt arrow ( ┴ ) or assigned parity −1 . Biologically , this corresponds to repression . Similarly , the parity can be assigned to the edges from u ( t ) to input variables based on the sign of ∂Fi/∂u . For simplicity , we restrict our attention to systems in which the signs of partial derivatives ∂Fi/∂u and ∂Fi/∂Xj are the same over the entire domain in which functions Xj ( t ) and u ( t ) take their values . Note that self-loops , i . e . edges that connect nodes to itself , are not included in such representation; thus the sign of ∂Fi/∂Xi is irrelevant . Fig 2 illustrates such graph for a particular system . Directed paths , i . e . sets of consecutive edges , of such graph describe how signal u ( t ) propagates through the dynamical system . If all the directed paths to the output variable node Xn ( t ) go through a node Xi ( t ) , then perturbations to these node that make it insensitive to the signal ( e . g . gene knockout that sets Xi ( t ) ) would also imply that Xn ( t ) is insensitive to the signal . For example , in Fig 2 all directed paths from signal to nodes X6 , X7 , X9 , X10 and X11 go through X5 . For each directed path in the graph consisting of multiple edges we can define a combined parity corresponding to a product of individual parities of edges from which each path consists . For example , on Fig 2 the directed paths from X1 to X4 to X5 and from X1 to X3 to X5 are positive whereas the directed path from X2 to X4 to X8 is negative . Note that directed paths can include full circles around indirect feedback loops , e . g . from X5 to X6 to X7 to X6 to X9 . With these definitions and notation in mind we are ready to formulate a general theorem that states a sufficient condition for the system output Xn ( t ) to be a monotonic function of time . For convenience , two alternative ( but mathematically equivalent ) formulations of the theorem are given–one for the case in which the input signal u ( t ) is known and another one for which internal node Xi ( t ) serves as a proxy for some unknown signal . Theorem . For a dynamical system given by Eq ( 2 ) which is initially in steady state ( i . e . F ( X ( 0 ) ;u ( 0 ) ) = 0 ) the response of the output Xn ( t ) will monotonically increase ( or decrease ) in time response to changes in the input u ( t ) if: or We note that if u ( t ) is a monotonically decreasing function simple change of variables ( e . g . u→ −u ) will result in monotonically increasing input and the theorem can still be used . For example , on Fig 2 the first formulation for the theorem allow us to conclude that monotonically increasing input u ( t ) will ensure monotonic increase of , for instance , X4 ( t ) , since both directed paths from u to X4 ( via X1 or X2 ) have positive parity . By repeating this reasoning for the remaining nodes , we can conclude that monotonic increase is ensured for X1 , X2 , X4 , X5 , X10 and X11 , monotonic decrease is ensured for X3 , but monotonicity cannot be guaranteed for X6 , X7 , X8 and X9 . On the other hand , if we do not know whether input signal u ( t ) is monotonic or in case an additional negative path in the network from u ( t ) to X5 is added , we may still use the second formulation to conclude that if X5 ( t ) is monotonic so will be X10 and X11 . Indeed , all the paths to X10 and X11 from input u ( t ) pass through X5 and all the paths from X5 to X10 and X11 have positive parity . The argument does not work for X9 due to a negative feedback loop between X6 and X7 ( a directed path that goes around this loop will have the opposite parity from the path that does not ) . One straightforward consequence of this theorem states that for any dynamical system in which a certain output variable Xn ( t ) behaves non-monotonically as a function of time in response to a monotonic signal , there must be at least two ( undirected ) paths between that node and input node u ( t ) with different parities , i . e . one with an odd number and another with an even number ( or zero ) of negative interactions . Such paths can only exist if the corresponding graph exhibits incoherent feedforward loops and/or negative feedback loops . This property can be very useful to identify that known biochemical network diagram is inconsistent with measured dynamics as we illustrate below . Applying the formulated results to the dynamics of icl1 transcription one can observe that the non-monotonic induction of icl1 ( Fig 1B ) is inconsistent with the network diagram proposed in Ref . [12] ( Fig 1A ) as the hypoxia signal only affects icl1 via sigE ( all the directed paths go through sigE ) , and there is no negative parity ( inhibiting ) path from sigE to icl1 . Nevertheless , the monotonic increase in sigE results in non-monotonic dynamics of icl1 . To further pinpoint the inconsistency , we have repeated the experiments of [12] and additionally measured the mRNA dynamics of all the intermediate species . The results of these measurements are shown in Fig 3A . By applying the monotonic systems theory to these new experimental results we can pinpoint at least two missing negative loops in the previously proposed network architecture ( Fig 1A ) . First , the monotonic increase in sigE mRNA leads to non-monotonic increase in its direct target , sigB , which increases between days 0 and 4 and then decreases on day 5 . We note that the difference between sigB mRNA in days 4 and 5 is small and statistical significance of the decrease is questionable ( p~0 . 1 ) . However if the decrease is real , there must be a negative loop in the network that posttranscriptionally regulates activity of σE or activates transcriptional repressor of sigB . However , regardless of this negative loop , another interaction is required to explain the observed icl1 dynamics . We note that icl1 starts to decrease after peaking at day 3 ( day 3 value is larger than that of day 0 and day 5 , p≤0 . 05 ) ; therefore , there must be at least one more negative loop acting downstream of sigB transcription . In fact , during the first 4 days of hypoxia , the dynamics of sigB is monotonic whereas the dynamics of lrpI , its transcriptional target , is not . Furthermore , all directed paths from the hypoxia signal to lrpI pass through sigB . Therefore , the formulated theorem predicts either a negative feedforward loop between sigB and lrpI or alternative signaling paths from hypoxia to lrpI transcription with a negative parity . In the subsequent sections we focused on uncovering this loop; our approach and workflow is illustrated in S1 Fig . The analysis above predicted existence of a negative loop that controls transcription of lrpI gene . However , no such loop could be found among known network interactions . One possible way to identify such a loop would be to find a transcriptional repressor of lrpI in the σB or σE regulons . We therefore examined these regulons for candidate genes with a DNA binding domain and transcriptional regulator function . One such gene is clgR ( Rv2745c ) , which encodes a transcription factor regulated by the MprAB-σE signaling system in response to various stressors , including redox stress , heat shock , and hypoxia [18–20] . ClgR has been reported to induce multiple chaperones and proteases [11 , 21] , including the Clp protease system , which can degrade misfolded proteins and is critical for mycobacterial survival under stress conditions [19] . Since our recent work [9] uncovered a complex network modulating its activity under cell envelope stress , we decided to examine its effect on gene expression in hypoxia . To this end , we used RT-qPCR to measure the dynamics of clgR expression in wild-type cells and examine the expression of sigB , icl1 and lrpI in a clgR deletion mutant ( ΔclgR ) strain . Deletion of clgR resulted in a sigB expression pattern that is similar to wild-type but displays about 50% reduction in peak expression ( compare Fig 3A and 3B ) . This is not surprising given that the ClgR-induced protease Clp is responsible for a degradation of the specific anti-σE factor , RseA [11] . Therefore , the genetic deletion of clgR reduces sigB transcription by breaking a positive feedback loop that controls σE activity . In contrast , expression peaks for icl1 and lrpI are increased and shifted to day 4 ( Fig 3 ) . We note that in the ΔclgR strain non-monotonic induction of lrpI follows non-monotonic induction of sigB and , therefore , the negative loop controlling lrpI transcription may no longer be active . The simplest way to reconcile these results is to hypothesize repression of lrpI by ClgR ( Fig 4A ) . Even though ClgR has only been shown to positively regulate its gene products , numerous examples exist of a transcription factor that can act both as activator and repressor of transcription ( e . g . , the Bacillus subtilis master sporulation regulator , Spo0A , which activates about 40 genes and inhibits 81 genes [22] ) . To test whether ClgR-mediated repression of lrpI ( Fig 4A ) can explain the observed dynamics , we first tested this hypothesis by constructing a mathematical model of the network . Given the complexity of clgR regulation [9] we included its transcriptional profile as an input to a model by taking the measured data and using an interpolation function to generate a continuous function that follows the observed dynamics ( Fig 3C ) . A similar strategy was used for sigB expression ( in both wild-type and ΔclgR strains ) –another input to the network ( Fig 3C ) . These inputs were used in a deterministic , ordinary differential equations model describing the dynamics of σB , ClgR , Icl1 and LrpI protein concentrations and algebraic equations for their respective mRNAs ( we assumed the degradation is fast and concentrations are in quasi-steady state ) . We then performed multidimensional parameter optimization to determine whether the measured dynamics of mRNA can be matched by those produced by the model . It is important to note that while we used parameter optimization as a feasibility check for various network structures , we understand that the data presented is not sufficient to restrict the parameters . Our simulation results ( Fig 4B ) showed that the proposed network topology can match the observed data for icl1 but not for lrpI . The failure of this model is due to a non-linear gain between lrpI and sigB in the data . Indeed , between days 0 and 3 sigB mRNA increases about 2-fold whereas lrpI mRNA increases over 3-fold . Given that sigma factors bind the core RNA polymerase as a monomer and function non-cooperatively , this result cannot be due to sigB mRNA increase . In fact , in steady state we expect σB protein concentration to scale linearly with sigB mRNA and the transcriptional activity of σ-factors is usually expected to scale sub-linearly ( hyperbolic , Michaelis-Menten-like expression ) with their concentration [23] . Therefore , we would expect that , in steady-state , fold-change of σB -activated mRNA would be lower then fold-change of sigB mRNA . Repression by ClgR would further suppress the fold-increase of lrpI . The arguments can be formalized and generalized to transient gains as well ( see Methods for a rigorous mathematical proof as an application of the corollary of the formulated theorem in S1 Text ) . Additionally , we demonstrated that even with large variations in the sigB and clgR input curves , lrpI always has a sub-linear amplification ( S2 Fig ) . To resolve this discrepancy , we need a non-linear amplification in the transfer function between sigB and lrpI . For example , this amplification can be explained if lrpI positively regulated its own transcription ( Fig 4C ) . Indeed , incorporating these interactions into our model leads to a good agreement between model predictions and experiments for both wild-type and ΔclgR strain ( Fig 4D ) . To experimentally test autoregulation of lrpI , we used an lrpI knock-out mutant , in which the lrpI open reading frame is interrupted by a transposon insertion . While the strain does not produce functional LrpI protein , it is still possible to quantify the expression of the truncated lrpI mRNA by using primers and probes mapping to lrpI sequences located upstream of the transposon insertion . In contrast to model predictions , the measured lrpI expression in ΔlrpI was not statistically significantly different compared to wild type for all time points ( Fig 4D ) . This result rules out lrpI autoregulation . Therefore , another factor must be responsible for the non-linear amplification between sigB and lrpI mRNA . We then investigated an alternative hypothesis , in which post-translational regulation of σB may lead to a non-linear relationship between sigB expression and σB activity . To test the hypothesis that σB activity is post-translationally regulated , we decided to examine the dynamics of another target . To this end we selected another gene in the σB regulon , ideR ( Rv2711 ) , to serve as a reporter for σB activity . IdeR is a global transcription factor that helps to maintain iron homeostasis and is essential for mycobacterial virulence [24] . Since no other transcriptional inputs have been found ( Fig 5A ) , we assume that ideR mRNA represents a surrogate of σB activity [25] . We therefore used RT-qPCR to quantify how ideR expression dynamically changes under hypoxia . Notably , the measured ideR dynamics is similar to that of lrpI in two important aspects: ( i ) the fold increase of ideR mRNA between days 0 and 3 exceeds that of sigB mRNA in the same period ( ideR has more than 3-fold increase; sigB has less than 2-fold increase ) ; ( ii ) ideR peaks at day 3 and decreases at day 4 despite the increase in sigB transcription ( Fig 5 ) . These results indicate that σB activity is post-translationally regulated and the missing negative loop must involve post-translational regulation steps . Analyzing alternative ways for post-translational downregulation of sigma-factor activity , we identified two theoretical possibilities . One is the downregulation of activity by sequestration of an active form ( e . g . by an anti-sigma factor ) , another is the downregulation of protein level via proteolytic degradation . Since we expect the negative loop to involve ClgR , we hypothesized that ClgR-activated induction of anti-σB or protease degrading σB can lead to the observed dynamics . Anti-sigma factors are ubiquitous across bacterial genera , and have been shown to be present in a diverse array of species [26] . Many of the known sigma factors in M . tuberculosis have corresponding anti-sigma factors; however , an anti-sigma factor corresponding to the mycobacterial σB has not yet been identified [27] . Nonetheless , we decided to consider the possibility of a hypothetical anti-σB . While we were unable to definitively exclude the presence of a novel anti-sigma factor B , the inability to fit the models of the various ( rather complex ) networks to the experimental data ( S3 Fig ) coupled with the fact that has been no indication of an anti σB factor in any mycobacterial species [7 , 27] led us to set aside this hypothesis . As previously mentioned , ClgR is known to induce multiple essential protease systems in M . tuberculosis , which regulate the activity of numerous proteins through selective degradation [11 , 19 , 21] . Thus , we hypothesized that one of these protease systems , e . g . Clp , may modulate post-translational σB activity by selectively degrading the sigma factor ( Fig 6A ) . Incorporation of these interactions into the model leads to simulated dynamics that are in good agreement with the experimental data under a set of physiologically relevant parameters ( Fig 6B and 6C; S4 Fig ) . Indeed , this model is able to replicate the lrpI , ideR and icl1 mRNA dynamics both in wild type and ΔclgR strains . We further demonstrated that the qualitative output dynamics of the model were very robust to variations in the sigB and clgR input curves ( S5 Fig ) . These results indicate that the introduction of the previously unknown interaction between Clp and σB is sufficient to explain all observed dynamics in the icl1 network .
The combination of traditional molecular genetics and novel high-throughput assays has generated a vast amount of information on the interactions that comprise biochemical networks inside living cells . For example , transcriptional regulatory networks can be obtained from gene-expression analyses such as qPCR , gene expression microarray technology , and RNA-seq , while DNA-protein interactions can be studied by chromatin immunoprecipitation-based ( ChIP ) measurements . However , regulatory networks are incomplete in even the best studied model systems [28 , 29] . At the same time , as more dynamical information about the responses of these networks to external perturbations is accumulated , mathematical models can use this data to pinpoint the inconsistencies between observed gene expression dynamics and presumed network topology , and to predict missing interactions [30] . In this context , dynamical systems theories that formulate necessary or sufficient conditions for a given network topology to lead to certain dynamical behaviors are especially useful as these conditions are often independent from detailed reaction mechanisms and kinetic parameters . Much research exists on the characterization of dynamics of bio-molecular networks by means of their topology . Among many such directions of work , one may mention: ( 1 ) the deep theory worked out by Feinberg in the early 1970s based on the idea of deficiency [31] , which has been applied to many fields , including T-cell kinetic proofreading models [32] or receptor-ligand pharmaceutical models [33] , and is still the subject of a major research effort [34 , 35]; ( 2 ) the use of graph-theoretic ideas based on Petri nets [36]; ( 3 ) the theories of cooperative and competitive systems [37]; and ( 4 ) methods of commutative algebra and algebraic geometry [38 , 39] . In this work , we formulated the necessary conditions for an output variable of a dynamical system to non-monotonically vary in time with changing inputs . In simplified terms , a non-monotonic response to monotonic stimulus requires the presence of an indirect negative feedback or an incoherent feedforward loop in the network graph . Even when the exact stimulus dynamics is not known , the comparison of the dynamics of internal components with the topology of the network graph can point to inconsistencies or missing loops . We applied these results in combination with mathematical modeling and subsequent experimental tests to a network controlling transient upregulation of icl1 , the gene encoding a glyoxylate shunt enzyme ( isocitrate lyase 1 ) , in response to hypoxia in M . tuberculosis . We found at least one inconsistency in the previously postulated network structure . Our results predict that there should be biochemical interactions that post-transcriptionally downregulate sigma factors σB , and possibly σE . Focusing on the former , we show that this downregulation can be explained by proteolytic degradation of σB protein by Clp protease system . Our model demonstrates that this interaction is consistent with all measurements for wild-type and genetically modified strains ( Fig 6 ) . As ClpP is essential for M . tuberculosis growth in vitro [40] , a ClpP knockout mutant could not be created . Thus , this is a major prediction of the model that will be tested in future studies . The predicted proteolytic degradation of σB is not unprecedented . In multiple bacterial species , stress-induced alternative sigma factors are proteolytically degraded by clp homologs . For example , RpoS ( σS ) is an enterobacterial sigma factor that is implicated in stationary-phase survival as well as virulence in pathogenic species [8 , 41–45] . The protease ClpXP has been implicated in the regulation of RpoS in several pathogens , including S . typhimurium [43 , 44] , and has been directly shown to degrade RpoS in E . coli [45] . It is known that the mycobacterial σB and RpoS share an evolutionary relationship [8] , which might suggest that proteolytic degradation by ClpP is conserved among some sigma factors . Additionally , ClpP has been shown to degrade σB of B . subtilis [46] . Despite sharing a common name , the σB of B . subtilis is actually more closely related to the mycobacterial σF [27]; however , mycobacterial σF and σB share a common evolutionary origin [27] , again possibly pointing to a potential conserved mechanism for the regulation of stress-induced sigma factors . What is the significance of the transient activation of icl1 in hypoxia and the predicted feedforward loops ? The hypoxic response can be considered a component of the metabolic adaptation of M . tuberculosis during infection , both because oxygen may become limiting inside the infected macrophage and because the lung tissue microenvironment becomes hypoxic as a granuloma develops [47–49] . Such adaptation ultimately leads to tubercle bacilli transitioning to a non-replicating/persistent state that is associated with latent infection . It may be argued that the stress-induced activation of adaptive metabolic pathways , such as those involving the activity of the icl1 product , is transient because the response to stress is followed by slow-down of the mycobacterial metabolic activity . Consequently , no sustained expression of metabolic enzymes such as Icl1 is required . Presumably , the transient surge in icl1 transcription at day 4 of hypoxia produces sufficient amounts of Icl1 protein until persistence is fully attained . Indeed , the observation that icl1 is induced transiently also during mouse lung infection supports a physiological role of this dynamics in pathogenesis [12] . If these hypotheses are correct , drugs blocking negative interactions responsible for non-monotonic dynamics could in principle destabilize transitions to latency or trigger reactivation . The theorem described in this work can be applied to a wide variety of biological systems to help understand the relationship between network topology and dynamics . For example , biochemical adaptation is a topic of wide interest , largely because of its ubiquity in biological systems—adaptation occurs when a step-up input into a biochemical network causes one or more downstream components to transiently increase but then return to the previously maintained steady state . Chemotaxis—the process where a cell moves in response to a chemical signal—is a well-known behavior that exhibits adaptive responses . This behavior is seen in a wide variety of organisms , including bacteria , neutrophils [50] and amoeba [50–52] . The mechanisms whereby these organisms achieve adaptation vary widely , but a well-studied case is chemotaxis in E . coli—given a step input of a chemoattractant , the rate of ‘tumbling’ will initially sharply decrease , but then return to nearly its original level [53] . Therefore , biochemical adaptation is a specific case where a monotonic input gives rise to a non-monotonic output . We may apply the described theorem to conclude that there must be either an incoherent feedforward loop or a negative feedback loop . Indeed , this observation is recapitulated by previous work by Ma et al . that used simulations and parameter sampling to show that there are only two types of networks that can produce robust biochemical adaptation–incoherent feedforward loops and negative feedback loops [54] . Thus , the theorem may help clarify structure-function relationships in well-characterized biochemical networks , as well as predict previously unforeseen inconsistencies in others .
The transcriptional regulation of both sigB and clgR is complicated and not very well understood [9 , 23] . To simplify the model , we treated sigB and clgR as inputs into the system by interpolating the time course of experimentally measured mRNA concentration of both species with a phenomenological function that followed the data trends . The interpolation gives a smooth , continuous function that can approximate species dynamics and can be fed directly into the model . As the wild-type sigB and clgR data appeared to demonstrate approximately adaptive dynamics , i . e . the time-point at day 5 is approximately the same as day 0 , the following pulse function was fitted to the data: mB ( t ) =1+ ( a1tna2n+tn ) ( 11+ ( ta3 ) m ) ( 3 ) mClgR ( t ) =1+ ( b1tnb2n+tn ) ( 11+ ( tb3 ) m ) ( 4 ) Here m , n and all an / bn are unknown parameters , and t is the independent variable ( time ) . The parameters n and m were fixed to 6 and 12 respectively , while the values ai and bi ( i = 1 , 2 , 3 ) were optimized to ensure best fit to the measured data . To this end , experimental data normalized to the value at day 0 , and non-linear least-square regression was performed using the MATLAB function fminsearch ( optimization toolbox ) to find the values of the unknown , free parameters . The data point at day 0 was replicated at day 1 for the sole purpose of interpolation as it was observed that relevant gene expression did not change significantly in our experimental set up from days 0–2 . Since the mean value of sigB mRNA in the ΔclgR strain decreased below its initial level on day 5 , a different form of pulse function was fitted to the data: mB ( t ) = ( c4+c1tnc2n+tn ) ( 11+ ( tc3 ) m ) ( 5 ) In the clgR mutant case , the parameters n and m were fixed to 6 and 18 respectively whereas the values of ci ( i = 1–4 ) were obtained by non-linear least-square regression as above . Fitted parameter values may be found in S1 Table . It is important to note that the form of the interpolating equations was chosen solely because of their adaptive behavior and ability to change quickly; the form of the equations does not have any biological significance . Therefore , we have also tested the robustness of the model to the precise form of the input by generating a family of input curves . In order to better understand both how uncertainty in the sigB and clgR expression measurements affected the dynamics of downstream nodes , we generated a family of input curves . Random data points were sampled from a normal distribution for each time point of both sigB and clgR , where the mean and standard deviation corresponded to the measured values . The point at day 0 was replicated to day 1 . 5 , and the point at day 5 was assumed to be near the final steady-state value and thus the point was copied to day 6 . These data points were interpolated with a shape-preserving cubic interpolation in MATLAB ( interp1 , method: ‘pchip’ ) . The interpolation curves were then each re-normalized to their initial value . We used the curves as in Figs S2 and S5 . The dynamics of protein concentrations in our model are given by ordinary differential equations that describe the kinetics of protein synthesis ( translation ) and first-order protein degradation . Here B ( σB; variable name B was used for ease of notation ) , LrpI , Icl1 , IdeR , and ClgR represent protein concentrations , and mx represents the corresponding mRNA concentrations . A description of all parameter symbols may be found in S2 Table . For B , concentration of σB protein , the dynamical equation is of the form: dBdt=bBmB−kdegBB ( 6 ) where mB is the concentration of sigB mRNA . In the same fashion , concentration of ClgR may be described by: dClgRdt=bClgRmClgR−kdegClgRClgR ( 7 ) The dynamics of LrpI and IdeR protein is similarly described: dLrpIdt=bLrpImLrpI−kdegLrpILrpI ( 8 ) As we do not have data describing Icl1 and IdeR protein dynamics , and Icl1/IdeR do not affect any other nodes in the network , we did not keep track of Icl1 and IdeR protein in our simulations . To solve these equations we also need equations for mRNA concentrations . As mRNAs usually have a significantly shorter half-life than proteins , we can employ the quasi-steady state approximation and describe mRNA concentration by algebraic equations as function of transcription regulators controlling their expression . For example , icl1 mRNA concentration ( micl1 ) is a function of σB and LrpI concentration: mIcl1=βicl1KILBnI+1+fILB ( B ) ( LrpI ) nIKILBnI+1+ ( B ) ( LrpI ) nI ( 9 ) Similarly , the concentration of both ideR and lrpI mRNA is solely a function of σB concentration . mLrpI=βLrpIKLB+fLBBKLB+B ( 10 ) mIdeR=βIdeRKRB+fRBBKRB+B ( 11 ) For the main model , used in Fig 6 , a ClgR-induced protease ( Clp ) was introduced , described by the following: mClp=βClpKCC+fCC ( ClgR ) nCCKCC+ ( ClgR ) nCC ( 12 ) dClpdt=bClpmClp−kdegClpClp ( 13 ) It was hypothesized that Clp degrades σB , so a new term was added to the equation for dynamics of σB solely for this model: dBdt=bBmB−kdegBB−kcat ( Clp ) ( B ) KM+B ( 14 ) The Michaelis-Menten formulation of enzyme kinetics was used here to represent the degradation of σB by Clp ( which assumes that the enzyme-substrate binding is fast ) . However , similar results were obtained when using the full ODE representation of enzyme kinetics , where C is the enzyme-substrate complex , formed between σB and Clp . The formulation is shown below: dBdt=bBmB−kdegBB−kf ( B ) ( Clp ) +krC ( 15 ) dClpdt=bClpmClp−kdegClpClp−kf ( B ) ( Clp ) +krC+kcatC ( 16 ) dCdt=kf ( B ) ( Clp ) −krC−kcatC−kdegCC ( 17 ) The direct repression of lrpI by ClgR with LrpI autoregulation ( Fig 4C ) is modeled as: mLrpI=βLrpI ( KLB+fLBBKLB+B ) ( KLLnL+fLLLrpInLKLLnL+LrpInL ) ( 11+ ( ClgRKLC ) nc ) ( 18 ) For the model without lrpI autoregulation ( Fig 4A and 4B ) , we use fLL = 1 and as a result: mLrpI=βLrpI ( KLB+fLBBKLB+B ) ( 11+ ( ClgRKLC ) nc ) ( 19 ) For the simulations shown on S3A Fig , we use Eqs ( 7–9 and 19 ) for ClgR , LrpI , IdeR , Icl1 and use an additional equation for a hypothetical anti-sigma factor ( A ) , adds new terms of binding and dissociation to σB dynamics to describe the sequestration of B with A to form a complex ( C2 ) : dBdt=bBmB−kdegBB−kfBA+krC2 ( 20 ) A is the ( free ) concentration hypothetical anti-sigma factor and AT is the total amount of anti-sigma factor ( bound + unbound ) . As we assume that the total amount of A ( AT ) is constant , it was treated a system parameter . The complex then follows the following dynamics: dC2dt=kfB∙A−krC2−kdegCC2 ( 21 ) At quasi-steady state , we use the following expression: A= ( kr+kdegC ) ATkr+kdegC+kfB ( 22 ) The case where A is regulated by ClgR was also explored ( refer to S3C Fig ) . Most equations remain the same , except the total amount of A is no longer treated a system parameter , and the pseudo-steady state approximation is not applied to C . The following describe the mRNA and protein of A: mA=βAKACnA+fAC ( ClgR ) nAKACnA+ ( ClgR ) nA ( 23 ) dAdt=bAmA−kdegAA ( 24 ) Similarly , the case where σB activated transcription of the hypothetical sigma factor was also investigated ( refer to S3B Fig ) ; simply replace ClgR with B and set nA to 1 in Eq 24 . Below we demonstrate that regardless of the parameter values and for any monotonically increasing mClgR ( t ) > 0 and mB ( t ) > 0 , the solution dynamical system consisting of differential Eqs ( 6–8 ) and algebraic Eq ( 19 ) starting from steady-state at t = 0 would be subject to the following condition: mB ( t ) mB ( 0 ) ≥mLrpI ( t ) mLrpI ( 0 ) ( 25 ) To prove this , we first note that with monotonically increasing mClgR ( t ) in Eq ( 7 ) we can conclude that ClgR ( t ) ≥ClgR ( 0 ) ( 26 ) Now consider alternative dynamical system for which m~ClgR ( t ) =m~ClgR ( 0 ) and consequently Clg~R ( t ) =Clg~R ( 0 ) . Here and below ~ denotes variables of an alternative system . Now from Eq ( 19 ) we can see that mLrpI ( t ) ≤m˜LrpI ( t ) andmLrpI ( 0 ) =m˜LrpI ( 0 ) ( 27 ) As the last term in Eq ( 19 ) is a decreasing function of ClgR and the rest of the terms are the same in original and alternative system . We also note that the alternative system no longer has negative loop between the input , mB ( t ) , and the output , m~LrpI ( t ) . Therefore , we can apply the result on the steady state gain ( S1 Text ) to conclude that m~LrpI ( t ) ≤βLrpI[KLB+fLBbBmB ( t ) /kdegBKLB+bBmB ( t ) /kdegB] ( 11+ ( ClgR ( 0 ) KLC ) nc ) ( 28 ) Here the right-hand side is i/o steady state response of the system deduced from Eqs ( 6 ) and ( 18 ) ( G ( u ( t ) in the notation of the S1 Text ( page 5 ) , with mB ( t ) = u ( t ) as an input ) . We note that the expression in the square brackets is a sublinear function of mB ( t ) and therefore KLB+fLBbBmB ( t ) /kdegBKLB+bBmB ( t ) /kdegB≤mB ( t ) mB ( 0 ) KLB+fLBbBmB ( 0 ) /kdegBKLB+bBmB ( 0 ) /kdegB ( 29 ) Now by combining these results we conclude mLrpI ( t ) mLrpI ( 0 ) ≤m∼LrpI ( t ) m∼LrpI ( 0 ) ≤mB ( t ) mB ( 0 ) ( 30 ) We note that for t = 3 days we have mB ( t ) mB ( 0 ) ∼2 and mLrpI ( t ) mLrpI ( 0 ) ∼3 contradicting this inequality . Thus the model without autoregulation of can never match the observed fold-change in mLrpI The formulated system of equations for each model was analyzed using a number of tools and functions in MATLAB 2013 ( a ) . Solution of the system of ODEs was obtained with ODE15s solver , as the parameter variation during optimization may cause system stiffness . Each case was run using 500 different initial parameter sets , by setting the initial parameters to a set of random numbers generated through the RandStream function in MATLAB , seeded with 'mt19937ar’ . All parameter optimization simulations were run on the Rice Shared Tightly-Integrated Cluster ( STIC ) . Parameter fitting was employed in order to test the compatibility of the proposed network topologies with the experimental data . If there is a set of parameters that allows the dynamical equations describing the system to sufficiently replicate the experimental data , then the network topology may be feasible . In order to fit system parameters to the data , the parameters were varied in order to attempt to minimize the deviation of the numerical solution of the dynamical system from the experimental data ( least squares ) . All fits used unweighted least squares , except the wild type ideR and icl1 data , which was weighted by the standard deviations of each data point because the data point at day 3 and 4 respectively had particularly large standard deviations . Parameter optimization was performed using particle swarm optimization ( pso ) , a metaheuristic , constrained optimization routine . The same implementation of pso for MATLAB was used for all aforementioned cases [55] . The pso algorithm was set to have a maximum generation number of 5000 and a population size equal to the number of free parameters . The parameter constraints are delineated in S3 Table , and the parameters corresponding to the fit on Fig 6 are shown in S4 Table . In order to avoid assumptions regarding the underlying distribution of the data , a Wilcoxon rank-sum test ( in MATLAB ) was used to evaluate statistical significance for all tests . A one-tailed test was employed to evaluate if the peak expression was larger than both the first and last data points for the dynamics of each gene product ( sigB , clgR , lrpI , icl1 , ideR ) . A two-tailed test was used to evaluate if there was a significant difference between lrpI expression in wt and ΔlrpI strains for all time points . M . tuberculosis mutants with deletion of sigE , sigB , or clgR , were previously reported [56–58] and a transposon-insertion mutants in gene rv0465c was obtained from the BEI repository [12 , 59] . The gene numbering of the M . tuberculosis genome is presented according to the system of Cole et al . [60] . Aerated and hypoxic cultures of M . tuberculosis were grown in Dubos Tween-albumin broth ( Becton Dickinson ) or Middlebrook ( MB ) 7H10 ( solid medium ) ( Difco ) supplemented with 0 . 05% Tween 80 , 0 . 2% glycerol , and 10% ADN ( 2% glucose , 5% bovine serum albumin [BSA; Sigma] , 0 . 15 M NaCl ) . Aerated liquid cultures of M . tuberculosis were grown in 25-ml tubes at 37°C with magnetic-bar stirring at 450 rpm . Hypoxic cultures of M . tuberculosis were grown as described below [12] . Bacilli growth was monitored by measuring optical density or enumeration of colony forming units . Aliquots of cultures were harvested at selected time points and processed for RNA extraction . We note that we observed no growth defects in the mutants , and the optical densities across the time-courses were nearly identical for all strains used ( S6 Fig ) . When M . tuberculosis cultures reached OD580 of 0 . 4 ( mid-log phase ) , they were diluted to an OD580 of 0 . 004 . Gradual oxygen depletion was achieved by incubating 17 ml-aliquots of diluted culture in 25-ml culture tubes containing a magnetic stirring bar . This design results in a ratio of headspace air to medium of 0 . 5 , in accordance with the classical method established by Wayne and Hayes [61] . RNA extraction and enumeration of bacterial transcripts were performed as described previously [12 , 62 , 63] . Briefly , total RNA was purified using TRI reagent ( Molecular Research Center , Cincinnati , OH ) according to the manufacturer's protocol . Reverse transcription was performed with random hexameric primers and ThermoScript reverse transcriptase ( Life technology , Carlsbad , CA ) . Transcripts were enumerated by real time PCR in a Stratagene Mx4000 thermal cycler ( Agilent Technologies ) , using gene-specific primers , and molecular beacons ( refer to S5 Table ) . Transcript numbers were normalized to 16S rRNA copy number of M . tuberculosis , as described previously [12 , 62] . To compare with simulations , all time course qRT-PCR data in the wild-type strain were normalized to the first data point ( time 0 ) , whereas all data from mutant strains were normalized to the first time point of the corresponding wild type strain . | Over the last several years mathematical modeling has become widely used to understand how biochemical systems respond to perturbations . In particular , dynamics of the response , i . e . the precise nature of how the responses changes with time , has become the focus of multiple studies . However , to this date only a few general rules that relate the dynamical responses with the structure of the underlying networks have been formulated . To this end , we ask which properties of the network allow systems to have a non-monotonic time-response ( first increasing and then decreasing ) to a monotonically increasing signal . We show that the networks displaying such responses must include indirect negative feedback or incoherent feedforward loop . Applying this result to the measured non-monotonic expression for glyoxylate shunt genes in Mycobacterium tuberculosis , a network known to be important to mycobacterial virulence , we show that the currently postulated network structure does not match the predictions of the theorem . Using a combination of mathematical modeling and follow-up experimental test we predict a novel incoherent loop in the network . This methodology has wide applications outside the specific network studied in this work—the theorem may potentially simplify the analysis of many biological systems . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"genetic",
"networks",
"enzymes",
"gene",
"regulation",
"messenger",
"rna",
"enzymology",
"optimization",
"systems",
"science",
"mathematics",
"network",
"analysis",
"hypoxia",
"genome",
"analysis",
"bacteria",
"computer",
"and",
"information",
"sciences",
"genomics",
"proteins",
"gene",
"expression",
"actinobacteria",
"dynamical",
"systems",
"biochemistry",
"rna",
"cell",
"biology",
"nucleic",
"acids",
"mycobacterium",
"tuberculosis",
"genetics",
"biology",
"and",
"life",
"sciences",
"proteases",
"physical",
"sciences",
"computational",
"biology",
"organisms"
] | 2016 | Non-monotonic Response to Monotonic Stimulus: Regulation of Glyoxylate Shunt Gene-Expression Dynamics in Mycobacterium tuberculosis |
We carried out whole genome resequencing of 127 chicken including red jungle fowl and multiple populations of commercial broilers and layers to perform a systematic screening of adaptive changes in modern chicken ( Gallus gallus domesticus ) . We uncovered >21 million high quality SNPs of which 34% are newly detected variants . This panel comprises >115 , 000 predicted amino-acid altering substitutions as well as 1 , 100 SNPs predicted to be stop-gain or -loss , several of which reach high frequencies . Signatures of selection were investigated both through analyses of fixation and differentiation to reveal selective sweeps that may have had prominent roles during domestication and breed development . Contrasting wild and domestic chicken we confirmed selection at the BCO2 and TSHR loci and identified 34 putative sweeps co-localized with ALX1 , KITLG , EPGR , IGF1 , DLK1 , JPT2 , CRAMP1 , and GLI3 , among others . Analysis of enrichment between groups of wild vs . commercials and broilers vs . layers revealed a further panel of candidate genes including CORIN , SKIV2L2 implicated in pigmentation and LEPR , MEGF10 and SPEF2 , suggestive of production-oriented selection . SNPs with marked allele frequency differences between wild and domestic chicken showed a highly significant deficiency in the proportion of amino-acid altering mutations ( P<2 . 5×10−6 ) . The results contribute to the understanding of major genetic changes that took place during the evolution of modern chickens and in poultry breeding .
The modern chicken ( Gallus gallus domesticus ) was domesticated from the red jungle fowl ( RJF ) [1] , but with some contributions from at least one other closely related species , the grey jungle fowl [2] . Domestic chicken segregate into several hundreds of distinct breeds distributed across the world . During the last century , the domestic chicken has been developed into a major food source by adapting the genome to specialized egg laying ( layers ) and fast-growing meat birds ( broilers ) whose productivity far exceeds their wild ancestor as well as the domestic chicken present only 100 years ago . Most modern commercial layers produce ~300 eggs in a year while the RJF usually lay a single clutch of 5–9 eggs per year . Modern broilers rapidly reach a body weight of 4–5 kg compared to the ~1 kg weight of an adult RJF male [3] . The commercial broiler and layer suppliers produce more than 70 billion birds annually to meet current worldwide consumer demands of more than 120 million tons of meat and over 1 . 2 trillion eggs [4] . The increasing productivity has been achieved through intensive directional selection on production traits over several tens of generations in purebred populations of limited effective population size followed by crossbreeding strategies in the generation of production animals . Maximizing yield however , has been at the price of reduced immunity and accompanied by a number of undesirable traits [5] . These negative effects may either be the result of pleiotropy of genes under selection for increased productivity , hitch-hiking of unfavourable alleles with the alleles under selection , or genetic drift . Understanding the nature of adaptive forces acting on the genome of commercial chicken provides insight into the complex relationship between production , disease and genes while opening up new directions for further improvement of this important farm animal , that is essential for global food security . The domestic chicken is an excellent model to investigate the genetics of adaptation , as it involves transformation of the ancestral red jungle fowl into a domesticated bird . Furthermore , parallel populations of broilers and layers were independently established from earlier multi-purpose populations by several breeding companies selecting for very similar breeding goals during the last hundred years . This scenario allows us to explore if the same alleles are responsible for the selection response in these parallel populations . In this study , we conducted a systematic comparison of genomic sequence variation from multiple populations of broilers and layers , versus each other and versus RJF to identify genes that underwent selection during domestication and the subsequent specialization of domestic chicken into broiler and layer lines . We report the discovery and characterization of over 21 million SNPs , 34% of which were not previously described . Analysis of selection provides a comprehensive list of candidate loci underlying domestication and/or changes in production-relevant traits . We further report a highly significant ( P<2 . 5×10−6 ) deficiency of amino-acid altering mutations among those showing strong genetic differentiation between RJF and commercial birds .
The bioinformatics analysis using the described criteria detected ∼26 . 3 million putative SNPs and INDELs . After filtering , over 21 million high-quality bi-allelic SNPs were retained that were either segregating or fixed for a non-reference within a population . The retained variants were distributed in the genome with an average density of 1 SNP every ~50 bases . About 34% of these SNPs ( n = 7 , 146 , 382 ) had not been reported before . The number of SNPs detected in each population varied between 7 . 6 and 17 . 4 million ( Table 1 ) . For the layer lines , the proportion of segregating variants was rather low , with an average of 57% of total variation , while the corresponding average for the broilers was 65% . RJFt alone carries 86% of all detected variants . These results show that layers have lost a considerable amount of the genetic diversity present in their wild ancestor as also indicated by the significantly lower levels of nucleotide diversity ( π ) in LRs ( 0 . 15–0 . 20% ) compared with that observed in RJFt ( 0 . 40%; Table 1 ) , although the possibility exists that the nucleotide diversity in RJFt is somewhat inflated if multiple subpopulations in northern Thailand was sampled . The low nucleotide diversity of RJFi ( 0 . 13% ) is presumably due to the fact that this population has been maintained as a small , closed breeding population for many years . The observed reduction in nucleotide diversity in the layer lines is mainly attributed to small number of founders and many generations of mating within closed lines of limited population size , but also partly due to the effect of linked selection . We detected 115 , 107 amino acid-altering SNPs of which 17% were predicted by SIFT to be evolutionary intolerant ( SIFT scores = 0 . 00–0 . 05 ) , 215 , 810 synonymous variants , 588 , 491 variants within untranslated regions and 1 , 100 stop-gain or -loss variants . An unknown fraction of these will have functional consequences . The comparison of the allele frequency profiles of wild and commercial populations reveals substantial differences ( Fig 1A; S2 Fig ) . In wild birds ( RJFt ) , the distribution of allele frequencies shows a marked overrepresentation of infrequent alleles which is consistent with the pattern observed for high-quality data in many other organisms including human and cattle populations [6 , 7] . In contrast , commercial populations , particularly layers ( S2 Fig ) , show a substantially smaller proportion of rare alleles that can be attributed to the smaller effective population size caused by recent selective breeding leading to loss of rare alleles . A subtle excess in the proportion of missense relative to synonymous mutations is evident among rare alleles , presumably caused by selection reducing the allele frequency of slightly deleterious mutations [6 , 8] . Fig 1B shows the distribution of population- and group-specific variants detected from individual sequencing only . Out of the >18 million variants detected in RJFt , as many as 4 . 4 million were unique to this population ( Fig 1B ) . This suggests loss of genetic diversity during domestication and breeding , although this might be partly due to genetic differences between the RJF birds used in this study and the ancestral population ( s ) of red jungle fowl that contributed to chicken domestication . We compared the distributions of population-specific SNPs among commercial and wild chicken to investigate differences in the frequency patterns ( Fig 1C ) . With the exception of the inbred RJFi population , the layer lines exhibit higher frequencies of population-specific alleles . This may be a consequence of a more narrow genetic basis and successive generations of selective breeding in commercial populations to enhance the frequency of favourable alleles . A good proportion of these loci are probably dragged to higher frequencies due to linkage with the target loci under selection [9] . Summary statistics of group-specific variants discovered exclusively in the layer and broiler lines are presented in supplementary Tables 1–4 . We performed a comprehensive analysis of genetic similarity based on genotypes from >21 million SNPs . As expected , individually sequenced birds from the same population clustered together ( Fig 1D; S3 Fig ) . The white ( WL ) and brown ( BL ) laying birds clustered distantly , although they are both layers , a result consistent with previous data [10] . Rhode White ( RWp ) is a layer breed developed by crossing white and brown layers and is located in the middle of the plot . The clusters of RJFs from Thailand and India were in fairly close proximity to one other . Broilers showed a strong clustering in the middle of the plot , probably due to the common ancestor of all , rooted back to the Cornish breed [11] . These results provide important background information for our attempts to identify loci under selection in the domestic populations . The extensive SNP data combined with annotation information for each single site enabled us to explore the genomic distribution of sequence polymorphisms showing strong genetic differentiation between wild and domestic chicken as well as between broilers and layers . We carried out enrichment analyses to identify categories of SNPs showing differentiation between groups of birds . The absolute allele frequency difference ( ΔAF ) was calculated for different categories of SNPs in four contrasts ( 1 ) RJFs vs . Coms , ( 2 ) BRs vs . LRs , ( 3 ) RJFs vs . BRs and ( 4 ) RJFs vs . LRs and these ΔAF-values were sorted into 10 bins of allele frequency ( ΔAF 0–0 . 1 , etc . ) to test for possible enrichment of variants in different annotation categories among SNPs showing strong differentiation . In all contrasts the great majority of SNPs showed a ΔAF<0 . 10 ( Figs 4 and S6 , S11–S14 Tables ) . This implies lack of differentiation between groups of birds at most loci , whereas a small percentage of variants , including those under selection showed highly significant differentiation . The intensity of adaptive and purifying selection varies across the genome according to the functional properties; as such intergenic sequences evolve relatively more freely than protein-coding sequences . We observed a marked decline in relative abundance of missense substitutions showing a steady decrease above ΔAF = 0 . 2 in all contrasts ( Fig 4 ) . SNPs with marked allele frequency differences ( ΔAF≥0 . 7 ) between wild and commercial chicken demonstrate a highly significant deficiency of missense mutations ( P<2 . 5×10−6 ) . We argue that this sharp decline in the proportion of differentiated missense substitutions represents purifying selection that reduces the frequency of slightly deleterious mutations affecting production and/or health . Thus , SNPs showing strong genetic differentiation between wild and domestic chickens are enriched for selectively neutral variants that have changed in frequency due to genetic drift as indicated by the enrichment of intergenic SNPs ( P<0 . 0001 ) among variants with ΔAF>0 . 7 . This result is in sharp contrast to recently reported data for the Atlantic herring where a similar analysis of high ΔAF SNPs showed a highly significant enrichment of missense mutations and other functionally important variants in a species with huge population size and a minimum amount of genetic drift [19] . The increase of log2 values for the contrast RJF vs . Coms and the flat curve for BRs vs . LRs ( Fig 4 ) indicate most likely that a fraction of the missense mutations has been under positive selection during domestication . Therefore , we decided to focus on the highly differentiated missense variants ( e . g . , ΔAF>0 . 70 ) , which were only 262 and 188 in the contrast ‘RJFs vs . Coms’ and ‘BRs vs . LRs’ , respectively . All strongly differentiated missense variants in all four contrasts are compiled in S15–S18 Tables . Within the list of high ΔAF SNPs we observed multiple missense variants occurring in the same gene . For example , the 262 missense substitutions with ΔAF ≥ 0 . 70 in the RJFs vs . Coms contrast occur in only 189 different genes and the corresponding figure for the contrast BRs vs . LRs is 188 missense substitutions in 150 genes . This result may reflect hitchhiking or possibly the evolution of alleles composed of multiple causal variants affecting the function of the same gene as previously documented in domestic animals [20] . We used the hypergeometric test of FUNC [21] to perform a gene ontology enrichment analysis based on the list of all genes embedding differentiated missense mutations and found no significant overrepresentation of any particular biological process . Nevertheless , we noted that some of these variants occur in genes affecting domestication or production-related traits ( Table 4 ) . However , as most genes have pleiotropic effects , selection may possibly act on other functional effects of these genes than those highlighted here . In the following sections , we highlight some results from these analyses . Evolution of pigmentation traits from wild to domestic type is one of the most striking changes during domestication [20] . Traits associated with visual appearance in domestic chicken have been artificially selected for aesthetic reasons and as a trademark in establishing distinct breeds . In the enrichment analysis of ‘RJFs vs . Coms’ , two of the missense mutations with the highest ΔAF occur in the CORIN ( AFRJFs = 0 . 09 and AFComs = 0 . 96 ) and in Ski2 Like RNA Helicase 2 ( SKIV2L2 , AFRJFs = 0 . 71 and AFComs = 0 . 00 ) genes . CORIN is a modifier of Agouti signalling protein ( ASIP ) in dermal papilla and its absence causes ASIP activity being prolonged leading to lighter coat color in mice [22] . SKIV2L2 regulates melanoblast proliferation during early stages of melanocyte regeneration [23] . Thus , both genes are involved in the pigmentation process . However , no genotype-phenotype association has yet been established for CORIN and SKIV2L2 in chicken . Among the top ΔAF alleles in the ‘RJFs vs . Coms’ contrast is the gene encoding sperm flagellar protein 2 ( SPEF2 , AFRJFs = 0 . 03 and AFComs = 0 . 82 ) . SPEF2 is implicated in feather development . In contrast to the modern chicken , jungle fowl use feathers for flight and thermoregulation , both of which are more crucial in wild birds than in commercial chicken maintained in a controlled environment . However , thermoinsulation must have been an important trait in domestic chicken in the past when birds were kept in unheated stables in cold climate . Furthermore , SPEF2 is a major candidate gene for chicken early- and late-feathering [24] , which is an economically important trait in the poultry industry since it can be used to sex chickens , and likely another reason for the differentiation of this mutation through linked selection . Two other notable mutations in this contrast overlapped the GLI Family Zinc Finger 3 ( GLI3 , AFRJFs = 0 . 03 and AFComs = 0 . 79 ) and the Kinesin Family Member 7 ( KIF7 , AFRJFs = 0 . 03 and AFComs = 0 . 82 ) genes , both involved in Sonic hedgehog ( Shh ) signaling pathway that controls the normal shaping of many tissues and organs during embryogenesis including limb and wing development [25 , 26] . Further genetic and functional studies of these allelic variants are necessary to verify their possible contribution to chicken domestication . Coding SNPs with ΔAF≥0 . 7 in the contrast between BRs vs . LRs also included interesting candidate mutations . For example , a missense mutation of extreme ΔAF ( AFBRs = 0 . 14 and AFLRs = 0 . 86 ) , occur in the Leptin receptor gene ( Table 4 ) , whose function in regulating feed intake and body weight is well documented in mammals [27 , 28] whereas the role of leptin-leptin receptor interaction for metabolic regulation in birds is not yet clear [29] . Another particularly interesting substitution in this contrast overlaps the multiple epidermal growth factor 10 gene ( MEGF10 , AFBRs = 0 . 82 and AFLRs = 0 . 00 ) on GGA8 , known to function as a myogenic regulator of satellite cells in skeletal muscle [30] . Mutations in MEGF10 have previously been shown to cause an unusual combination of dystrophic and myopathic features leading to the weak muscles in humans [30 , 31] , suggesting that the mutation reported here may affect muscle growth in broilers . The fact that different broiler lines have a high frequency of the variant allele at this locus is consistent with this suggestion . Other notable mutations in this contrast were found in the IGSF10 gene implicated in age at puberty [32] and PLEKHM1 with a suggested role in osteoporosis [33] . This paper reports the discovery and characterization of over 20 million SNPs from the chicken genome with the goal to delineate those with potential functional consequences—either having adaptive advantages or deleterious effects . To our knowledge , this is so far the largest study of its kind in chicken as a large number of individuals have been sequenced and a large number of sequence variants were detected . As many as 34% ( n = 7 , 146 , 383 ) of the SNPs had not been reported before . The results revealed a subtle differentiation between wild and modern chicken at most loci , whereas a small percentage of loci showed strong differentiation . Analysis of selection provided a comprehensive list of several tens of independent loci that are likely to have contributed to domestication or improving production . We confirmed strong differentiation between red jungle fowl and domestic chickens at the previously reported BCO2 and TSHR loci . We identified 34 putative selective sweeps co-localized with , among others , KITLG , ALX1 , IGF1 , DLK1 , JPT2 and CRAMP1 . Single SNP contrasts between groups of birds revealed several highly differentiated coding variants , in genes such as CORIN and SKIV2L2 involved in pigmentation and LEPR , MEGF10 and SPEF2 possibly affecting traits relevant for animal production . SNPs with marked allele frequency differences between wild and domestic chicken showed a highly significant deficiency of the proportion of missense mutations ( P<2 . 5×10−6 ) .
Samples were either taken from a DNA bank established at Friedrich-Loeffler-Institut during the EC project AVIANDIV ( 1998–2000; EC Contract No . BIO4-CT98-0342 , https://aviandiv . fli . de ) or as part of the SYNBREED project ( 2009–2014 , Funding ID: 0315526; http://www . synbreed . tum . de/ ) where sampling was done in strict accordance to the German Animal Welfare regulations ( 33 . 9-42502-05-10A064 ) and with written consent of the animal owners . Three groups of birds were included in the study ( 1 ) red jungle fowls ( Gallus gallus gallus , RJFs ) , ( 2 ) broilers ( BRs ) and ( 3 ) layers ( LRs ) ( Table 1 ) . The RJFs were sampled from two geographical regions , Thailand ( RJFt ) and India ( RJFi ) . The RJFt consisted of 25 DNA samples collected within a European collaborative research project AVIANDIV ( https://aviandiv . fli . de/ ) . RJFt was randomly down-sampled from ~150 RJFs caught in northern Thailand in 1997 and maintained since with random mating over four flocks; given the place and date , the RJFt samples likely have seen some contamination from domestic or feral populations prior to collection [34] . The DNA samples from RJFt were collected in 1999 . For further information on the behaviour and morphology of these birds we refer to the AVIANDIV project webpage . The RJFi population involved 10 individuals of the Richardson line , originating from RJF caught in India in the 1960´s . This population has been extensively studied [35–37] , and appears to have been established from a wild population prior to major genetic contamination of red jungle fowl populations , such that it may represent a unique RJF line that is at least largely free of influence from domestic stocks . The second and third group of birds represent commercial chicken , comprising three broiler and three layer populations , respectively . The broilers ( BRs ) were represented by 20 DNA samples of each of two lines ( BRA and BRB ) established independently and previously collected as part of the AVIANDIV project . BRA was a sire line belonging to the company Indian River International ( Texas ) established in 1980 and closed since with a breeding population size of >10 , 000 birds . BRB was another sire line originally from France , developed in 1970 with a breeding population size varying between 10 , 000 to 70 , 000 . The broiler group further involved a pooled sample of 25 birds from AVIANDIV’s broiler sire line D , hereafter denoted BRpD . This is a sire line originally from UK , established in 1974 and closed since with unknown population size . In the layer group ( LRs ) , data from 25 birds each from purebred white ( WL ) and brown ( BL ) egg laying populations , sequenced in the frame of the SYNBREED project ( http://www . synbreed . tum . de/index . php ? id=2 ) , were included . WL and BL birds represent parental lines of the LOHMANN Tierzucht GmbH that are originally established from White Leghorn and Rhode Island Red , respectively . Moreover , we used pooled sequence data of 48 birds from Rhode Island White ( RWp ) , a crossbred layer population collected by the AVIANDIV project . Sequencing libraries of 300–500 bp fragments were constructed for each individual sample using Illumina Nextera Library preparation kits . Sequencing of RJFt , BRA and BRB was conducted using an Illumina HiSeq 2500 machine and 2x126 bp paired-end reads were generated . RJFi , WL and BL along with the three DNA pools ( RWp , BRpB and BRpD ) were sequenced with 2x101 bp paired-end reads ( see Table 1 ) . All reads were mapped against the reference genome assembly Galgal5 [38] using the Burrows-Wheeler aligner ( bwa-0 . 7 . 12 ) [39] . Duplicate reads were masked during pre-processing using the Picard tool set ( version 2 . 0 . 1 ) . We identified SNPs following the recommendations of best practices workflow for variant discovery analysis using GATK [40] . Briefly , after recalibrating for base quality scores , BAM files were fed into the GATK-HaplotypeCaller tool which is capable of calling SNPs and INDELs simultaneously via local de-novo assembly of haplotypes in a region . After generating 127 GVCF files for individual and pooled samples , they were called simultaneously using the GenotypeGVCFs module . Raw vcf files were then filtered and used for downstream analyses . S1 Fig presents a summary of SNPs called based on different sequencing parameters . The number of detected variants was 26 , 290 , 203 which included 3 , 442 , 027 INDELs and 1 , 024 , 944 multi-allelic sites . Raw vcf files from both individuals and pools were filtered primarily based on the following parameters . Variants were removed with QualByDepth ( QD ) < 4 . 0 , 300 > depth > 2200 , Quality < 30 , mapping quality ( MQ ) < 40 . 0 , MQRankSum < -10 , ReadPosRankSum < -7 . 0 , Fisher Strand > 60 . 0 , ReadPosRankSum > 7 , BaseQRankSum < -6 , BaseQRankSum > 6" . Cluster Size and ClusterWindowSize were set to 4 and 10 , respectively . For the subsequent analyses we used only bi-allelic SNPs on autosomes and chromosomes W and Z . In total , 21 , 190 , 795 SNPs were retained for downstream analysis . The R packages SNPRelate and gdsfmt [41] were used for principal component analysis of relatedness using identity-by-descent measures estimated from all SNPs . SnpEff ( v . 3 . 4 ) [42] was used to annotate variants according to their functional categorization which included the following categories 5 kb up- and down-stream of a gene , intergenic , missense , synonymous , intronic , 3' untranslated regions , 5' untranslated regions , stop gain and stop loss . Variants in the up- and down-stream regions and in the 3' UTR , 5' UTR regions were merged into the single categories . The online tool Ensemble Variant Effect Predictor ( VEP , webpage: http://www . ensembl . org/info/docs/tools/vep/index . html ) ) [43] , was used to predict SIFT-scores for amino-acid altering substitutions . The enrichment analysis was conducted as previously described in [44] for four contrasts ( 1 ) RJFs vs . commercial and ( 2 ) BRs vs . LRs , ( 3 ) RJFs vs . BRs and ( 4 ) RJFs vs . LRs . First we estimated the allele frequency ( AF ) of each SNP based on the proportion of high-quality reads supporting the non-reference allele . To ensure an unbiased estimation of AF several filters were employed to remove low quality SNPs and uncertain genotypes . In individually sequenced populations , loci with genotype quality < 20 were set to no . call and allele frequencies were estimated only for sites with >50% of the individuals genotyped . Because of low coverage , we treated the population RJFi as a pool in this analysis . In all pools SNPs with allelic depth <50% of mean coverage were set to no . call . Then , for each contrast , allele frequencies of intra-group populations were averaged and used to estimate the absolute value of allele frequency difference ( ΔAF ) for every single variant . The SNPs were then sorted into different bins of ΔAF ( e . g . , 0–0 . 1 , >0 . 1–0 . 2 , etc . ) representing the allele frequency difference between populations . The expected number of SNPs for each category in each bin was calculated as p ( category ) X n ( bin ) , where p ( category ) is the proportion of a specific SNP category in the entire genome and n ( bin ) is the total number of SNPs in a given bin . Finally , log2 fold changes of the observed SNP count for each category in each bin were compared against the expected SNP count and statistical significance of the deviations from the expected values was tested with a standard χ2 test . Evidence of positive selection was investigated in two steps . First , we explored differentiation of loci between the following combinations of populations . ( 1 ) RJFs vs . Commercials , ( 2 ) BRs vs . LRs , ( 3 ) RJFs vs . BRs and ( 4 ) RJFs vs . LRs . We estimated FST [45] for each of these four contrasts . To reduce locus-to-locus variation in the inference of selection we averaged single SNP values for sliding windows of 40 kb with 20 kb overlap across chicken chromosomes . Window-based FST values were then normalized and windows in the outlier tail ZFST > 6 were identified as selection candidates for domestication and genetic improvement in commercial populations . In the second step , we searched the genome for regions with high degrees of fixation . To this purpose , the nucleotide diversity ( Pi ) was compared between RJF and commercial birds as a signature of selection during domestication . Different window sizes were tested but did not change the consistent picture of the signals . A window size of 40 kb was selected in accordance to the differentiation analysis . The Pi values were then normalized . Analysis of fixation involved six populations for which individually sequenced data were available . As such , nucleotide diversity was estimated for RJFs ( two red jungle fowl populations ) , commercials ( four commercial lines ) , broilers ( the two commercial broiler lines , BRA and BRB ) , LRs ( two layer populations , BL and WL ) and ALL ( all six populations of RJFs and commercials ) . Gene ontology enrichment analyses , contrasting differentiated genes against a genomic background gene set , were performed using the hypergeometric test of FUNC [21] . | Domestic chickens ( Gallus gallus domesticus ) provide a critical resource for animal proteins for human nutrition worldwide . Chickens were primarily domesticated from the red jungle fowl ( Gallus gallus gallus ) , a bird that still runs wild in most of Southeast Asia . Human driven selection during domestication and subsequent specialization into meat type ( broilers ) and egg layer ( layers ) birds has left detectable signatures of selection within the genome of modern chicken . In this study , we performed whole genome sequencing of 127 chicken including the red jungle fowl and multiple populations of commercial broilers and layers to perform a systematic screening of adaptive changes in modern chicken . Analysis of selection provided a comprehensive list of several tens of independent loci that are likely to have contributed to domestication or improving production . SNP by SNP comparison of allele frequency between groups of wild and domestic chicken showed a highly significant deficiency of the proportion of amino acid altering mutations . This implies that commercial birds have undergone purifying selection reducing the frequency of deleterious variants . | [
"Abstract",
"Introduction",
"Results",
"and",
"discussion",
"Methods"
] | [
"animal",
"types",
"ecology",
"and",
"environmental",
"sciences",
"animal",
"genetics",
"domestic",
"animals",
"chickens",
"vertebrates",
"social",
"sciences",
"animals",
"animal",
"behavior",
"deserts",
"bird",
"genetics",
"zoology",
"ecological",
"metrics",
"birds",
"ecosystems",
"behavior",
"gamefowl",
"fowl",
"species",
"diversity",
"animal",
"genomics",
"poultry",
"bird",
"genomics",
"psychology",
"eukaryota",
"ecology",
"animal",
"sexual",
"behavior",
"genetics",
"biology",
"and",
"life",
"sciences",
"genomics",
"amniotes",
"organisms",
"terrestrial",
"environments"
] | 2019 | Genetics of adaptation in modern chicken |
RNA interference ( RNAi ) represents the only method currently available for manipulating gene-specific expression in Schistosoma spp . , although application of this technology as a functional genomic profiling tool has yet to be explored . In the present study 32 genes , including antioxidants , transcription factors , cell signaling molecules and metabolic enzymes , were selected to determine if gene knockdown by RNAi was associated with morphologically definable phenotypic changes in early intramolluscan larval development . Transcript selection was based on their high expression in in vitro cultured S . mansoni primary sporocysts and/or their potential involvement in developmental processes . Miracidia were allowed to transform to sporocysts in the presence of synthesized double-stranded RNAs ( dsRNAs ) and cultivated for 7 days , during which time developing larvae were closely observed for phenotypic changes including failure/delay in transformation , loss of motility , altered growth and death . Of the phenotypes evaluated , only one was consistently detected; namely a reduction in sporocyst size based on length measurements . The size-reducing phenotype was observed in 11 of the 33 ( 33% ) dsRNA treatment groups , and of these 11 phenotype-associated genes ( superoxide dismutase , Smad1 , RHO2 , Smad2 , Cav2A , ring box , GST26 , calcineurin B , Smad4 , lactate dehydrogenase and EF1α ) , only 6 demonstrated a significant and consistent knockdown of specific transcript expression . Unexpectedly one phenotype-linked gene , superoxide dismutase ( SOD ) , was highly induced ( ∼1600-fold ) upon dsRNA exposure . Variation in dsRNA-mediated silencing effects also was evident in the group of sporocysts that lacked any definable phenotype . Out of 22 nonphenotype-expressing dsRNA treatments ( myosin , PKCB , HEXBP , calcium channel , Sma2 , RHO1 , PKC receptor , DHHC , PepcK , calreticulin , calpain , Smeg , 14 . 3 . 3 , K5 , SPO1 , SmZF1 , fibrillarin , GST28 , GPx , TPx1 , TPx2 and TPx2/TPx1 ) , 12 were assessed for the transcript levels . Of those , 6 genes exhibited consistent reductions in steady-state transcript levels , while expression level for the rest remained unchanged . Results demonstrate that the efficacy of dsRNA-treatment in producing consistent phenotypic changes and/or altered gene expression levels in S . mansoni sporocysts is highly dependent on the selected gene ( or the specific dsRNA sequence used ) and the timing of evaluation after treatment . Although RNAi holds great promise as a functional genomics tool for larval schistosomes , our finding of potential off-target or nonspecific effects of some dsRNA treatments and variable efficiencies in specific gene knockdown indicate a critical need for gene-specific testing and optimization as an essential part of experimental design , execution and data interpretation .
Digenetic trematodes ( parasitic flatworms ) of the genus Schistosoma infect more than 200 million people in over 70 developing countries [1] , with an additional 770 million people worldwide at risk of becoming infected [2] . As causative agents of chronic , often severe morbidity and responsibility for an estimated 280 , 000 death per year in Africa alone [3] , schistosomiasis ranks as one of the most important of neglected tropical diseases [4] . Although significant research effort and funding have been dedicated to the treatment and control of schistosomiasis , including sanitary measures , suppression of the snail intermediate host , and chemotherapeutic interventions , there has been little change in the overall disease prevalence [5] . Progress in vaccine development has been very slow , and although several antigens , some of which are currently under clinical trial , have shown limited promise in rodent and primate challenge experiments , prospects are not good for an effective , highly protective vaccine in the foreseeable future [6] , [7] . Clearly there continues to be a pressing need for new strategies to break the cycle of schistosome transmission to the human population [8]–[10] . In view of the limited options available for controlling schistosomiasis in both the human host and snail vector , it is important that research focus on obtaining information that can be translated into new tools for parasite control . To that aim , genomic , transcriptomic and proteomic approaches offer strong possibilities to discover new potential targets for vaccines , develop new drug candidates , and provide a better understanding of basic molecular mechanisms underlying host-parasite interactions . The S . mansoni Genome Project and data generated by various gene discovery efforts using expressed sequence tags ( ESTs ) and serial analysis of gene expression ( SAGE ) , have resulted in a massive amount of gene sequence and expression information [11]–[19] . However , without reliable reverse or forward genetics methods , this vast amount of data cannot be placed into any functional context that can then be used to determine the value or importance of specific genes as targets for disease control . Unlike the parasitic nematodes , which have benefited from reverse genetic methods developed in the model free-living worm Caenorhabditis elegans [20] , [21] , no analogous model system is available for schistosomes . This has further delayed the application of new genomic technologies to problems related to disease control and drugs development [22] . However , despite this lack of a C . elegans-type model for parasitic flatworms , important advances have been made in trematode transgenesis with the introduction and transient expression of various reporter constructs in schistosomes [22] , [23] and fasciolids [24] , although these approaches did not permit the functional assessment of specifically introduced genes . With the first demonstrations of gene expression knockdown by RNA interference ( RNAi ) in the mammalian [25] and snail [26] stages of S . mansoni , this reverse-genetics approach has now been applied to a limited number of genes expressed in primary ( = mother ) sporocysts [27] , schistosomula [28] , [29] and adults [30] . In a recent review of RNAi in parasitic helminths , however , Geldhof and collaborators [31] admonish researchers for , at times , providing insufficient data that more firmly connects RNAi-mediated gene expression changes with specific phenotypes , and/or reporting only on genes that are susceptible to double-stranded ( ds ) RNA-mediated knockdown . Therefore , in order to gain a broader profile of RNAi efficacy in schistosomes , in the current study we performed an in vitro phenotypic screening of 32 genes known to be expressed in primary sporocysts of S . mansoni . These genes covered a variety of functional categories including antioxidants , transcription factors , cell signaling molecules and metabolic enzymes . Out of the 32 genes ( comprising 33 dsRNA treatments ) targeted for silencing , one-third ( 11 genes ) exhibited a dsRNA treatment-associated phenotype that consisted of a reduction in sporocyst size ( i . e . , larval length ) . Interestingly of the 11 phenotype-yielding genes , only 7 demonstrated a significant and consistent alteration in transcript expression after the 7-day treatment period , although time-course experiments suggest that transient gene knockdown during earlier times of exposure may , in part , account for the observed phenotype .
All experiments were performed using the NMRI strain of S . mansoni . Eggs were obtained from 7–8 weeks infected mouse livers . After hatching in an artificial “pond water” [32] containing antibiotics ( 50 µg/mL streptomycin and 60 µg/mL penicillin ) , miracidia were immobilized on ice for 15 min , washed twice in cold pond water by centrifugation ( 1 min , 700×g ) and gently resuspended in Chernin's balanced saline solution ( CBSS ) [33] , supplemented with glucose and trehalose ( 1 g/L each ) , streptomycin ( 50 µg/mL ) and penicillin ( 60 µg/mL ) [34] . Larvae were then counted and distributed into either 48- or 96-well polystyrene tissue culture plates ( Costar , Corning Incorporated , NY ) at concentrations of ∼6000 and 500 miracidia/well , respectively , depending on the experiments being performed . The general procedure used in all RNAi experiments involve treatment of miracidia starting at day 0 in culture with specific dsRNAs or control media containing irrelevant dsRNAs or medium alone for 7 days followed by assessment of an RNAi-type effect [26] . Details of dsRNA preparation and experimental designs are presented below . All research protocols involving mice used in the course of this study were reviewed and approved by the Institutional Animal Care and Use Committee ( IACUC ) at the University of Wisconsin-Madison under assurance no . A3368-01 . In the present study , a total of 32 genes were selected for quantitative and qualitative assessment . Twenty-three of these genes were chosen based on their abundant expression in in vitro cultured S . mansoni miracidia and/or primary sporocysts , using the SAGE database OrganismDB [18] http://gmod . mbl . edu/perl/site/s_mansoniest]: calcineurin B ( AJ276885 . 1 ) , lactate dehydrogenase ( LDH; U87629 . 1 . ) , Smad4 ( AY371484 . 1 . ) , Smad2 ( AF232025 . 1 ) , Smad1 ( AF215933 . 1 ) , 14 . 3 . 3 ( U24281 . 1 ) , epidermal growth factor receptor ( Scmeg; M86399 . 1 ) , phosphoenolpyruvate carboxykinase ( PepCK; AF120929 . 1 ) , calpain ( M74233 . 1 ) , hexamer-binding protein ( HEXBP; putative , organismDB:Tag623 ) , fibrillarin ( putative , OrganismDB: Tag 428 ) , elongation factor 1α ( EF1α; Y08487 . 1 ) , Rho 1 GTPase ( Rho1; AY158212 . 1 ) , Rho 2 GTPase ( Rho2; AY158214 . 1 ) , calcium ATPase 2 , ( Sma2; AF074400 . 1 ) , SPO1 ( AF109181 ) , protein kinase Cß ( PKCß; AY337620 . 1 ) , protein kinase C receptor ( PKC receptor; AF422164 . 1 ) , zinc finger DHHC domain ( DHHC; putative , OrganismDB: Tag 1180 ) , myosin-light chain ( AF071011 . 1 ) , calreticulin ( L24159 . 1 ) , high voltage-activated calcium channel subunit α ( Cav2A; AF361883 . 1 ) and high voltage-activated calcium channel ß-subunit 2 ( calcium channel; AY277532 . 1 ) . The remaining 9 genes were chosen for their predicted putative functions in the parasite ( antioxidants , transcription factors ) and ongoing characterization by our group: glycoprotein K5 ( AY903301 . 1 ) , zinc finger 1 ( SmZF1; AF316827 . 1 . ) , ring box ( SmRbx; DQ466078 . 1 . ) , glutathione peroxidase ( GPx; M86510 . 1 ) , thioredoxin peroxidase 1 ( TPx-1; AF121199 . 1 ) , thioredoxin peroxidase 2 ( TPx-2; AF157561 . 1 ) , superoxide dismutase ( SOD; M27529 . 1 ) , 26 kDa glutathione-S-transferase ( GST26; M73624 . 1 ) , and 28 kDa glutathione-S-transferase ( GST28; S71584 . 1 ) . TPx-1 and TPx-2 were used in combination to simultaneously silence both thioredoxin peroxidases . All of the above sequences are available from GenBank ( http://www . ncbi . nlm . nih . gov/Genbank ) or OrganismDB as indicated above . T7 promoter-tagged specific PCR primers were designed to amplify ∼500 base pair ( bp ) products for each of the targeted genes ( Dataset S1; Table 1 ) . A 500-bp green fluorescent protein ( GFP ) gene segment also was synthesized from the vector pAcGFP ( Clontech , Mountain View , CA ) to serve as a nonspecific dsRNA treatment-control . Following amplification , PCR products were separated on 1% agarose gels and purified using QIAquick gel extraction kit ( Qiagen , Valencia , CA ) , following the manufacturer's protocol . Each PCR product was sequenced and their sequences verified using the Basic Local Alignment Search Tool ( BLASTn , National Center for Biotechnology Information , NCBI ) . Double-stranded RNAs were synthesized from isolated sporocyst cDNA using T7 RiboMAX Express RNAi Kit ( Promega , Madison , WI ) , according to procedures outlined by the manufacturer . Briefly , dsRNAs synthesis reactions were allowed to incubate for 16 hr at 37°C prior to DNAse treatment . DsRNA products were then phenol/chloroform-extracted and purified by precipitation with isopropanol . The purified products were resuspended in diethylpyrocarbonate ( DEPC ) -treated water , quantified by measurement at OD260 and their integrity verified by 1% agarose gel electrophoresis . Samples were stored at −20°C until further use . Effects of dsRNA treatment on in vitro cultured S . mansoni larvae were performed in 96-well culture plates ( Costar ) in which approximately 500 miracidia were added to wells containing 50 nM of specific or control green fluorescent protein ( GFP ) dsRNA diluted in 200 µL CBSS or medium lacking any dsRNA ( no dsRNA control ) . Cultures were maintained at 26°C for 4 days , after which time an additional 10 nM of dsRNA was added to each well due to possible RNA degradation in culture [35] , followed by incubation for 3 more days . Over the 7 days culture period , sporocysts were monitored for the following phenotypes: failure/delay in transformation , loss of motility , tegumental lysis and granulation ( lethality ) and changes in larval growth . Parasite viability and morphological changes were monitored daily using a Nikon Eclipse TE 300 inverted epifluorescent microscope ( Nikon Instrument Inc . , Melville , NY ) . In addition digital images of live treated and control parasites were captured using a CoolPix EZ digital camera ( Nikon Instruments Inc . ) throughout the 7-day incubation period , allowing more detailed observations of larval morphology and to quantify sporocyst growth ( length measurements ) in treated vs . control larvae at day 7 . Length measurements from captured images were obtained and analyzed by Metamorph software version 7 . 0 ( Meta Imaging series , Molecular Devices , Sunnyvale , CA ) . Sporocysts exhibiting tegumental lysis or loss of surface/somatic integrity were excluded from measurements . Larval growth datasets for each experimental replicate were statistically analyzed using the Mann-Whitney U-test ( Wilcoxon-Sum of Ranks test ) with significance set at P≤0 . 05 . All treatments were performed in duplicate wells , and the experiment was independently replicated a minimum of 3 times on miracidia isolated from different batches of infected mouse livers . In addition , to verify dsRNA uptake by sporocysts we labeled Smad4 , lactate dehydrogenase and GFP ( specificity control ) dsRNAs with rhodamine using the Label IT kit ( CX-Rhodamine Labeling Kit; Mirus , Madison , WI ) , according to manufacturer's recommendations . Miracidia were in vitro-transformed to sporocysts in CBSS containing 50 nM labeled dsRNAs and subjected to epifluorescence photomicrography after 7 days of incubation . In order to demonstrate an association between phenotype and transcript expression real-time quantitative PCR ( q-PCR ) was used to determine steady-state transcript levels in specific dsRNA-treated sporocysts . In these experiments ∼6000 miracidia were distributed into a 48-well plate ( Costar ) and treated with 50 nM dsRNA diluted in CBSS ( 500 µL/well ) . Cultures were maintained at 26°C for 2 , 4 or 7 days prior RNA extraction and isolation . Cultures maintained for 7 days were supplemented with 10 nM dsRNA at day 4 . Sporocysts were extensively washed with CBSS in order to eliminate unabsorbed dsRNAs and shed ciliary epidermal plates , followed by extraction in Trizol reagent ( Invitrogen , Carlsbad , CA ) to isolate both total RNA and protein fractions from cultured larvae . The protein pellet was dissolved in the protein solubilization buffer ( 3 M Urea , 2% CHAPS , 40 mM Tris ) for use in Western blot analyses ( see below ) , while the isolated RNA fraction was resuspended in DEPC-treated water and subjected to DNAse treatment using the DNA-Free kit ( Ambion , Austin , TX ) to eliminate any contaminating genomic DNA . RNA samples were quantified and their purity assessed on a Nanodrop Spectometer ND-1000 ( NanoDrop Technologies , Inc . , Wilmington , DE ) . Quantitative PCR analysis was used to compare steady-state transcript levels between specific dsRNA-treated sporocysts and control treatments ( GFP-dsRNA ) . To accomplish this 0 . 5 to 1 µg total RNA , derived from at least three different extractions , was used to synthesize cDNA using Superscript III cDNA Synthesis kit ( Invitrogen , Carlsbad , CA ) following the manufacturer's protocol . The q-PCR reaction mixtures consisting of 2 . 5 µL of cDNA , 12 . 5 µL of Sybr Green PCR Master Mix ( Applied Biosystems , Foster City , CA ) , 10 µL of 600 or 900 nM primers ( determined after primer concentration optimization ) , were added to 96-Well Optical Reaction Plates ( ABI PRISM , Applied Biosystems ) for amplification and quantification in a AB7500 Real Time PCR System ( Applied Biosystems ) . In order to avoid the possibility of false amplification of the originally applied dsRNA , specific pairs of primers were designed outside of the region used to synthesize the original interfering dsRNA products ( Dataset S1; Table 2 ) . In addition to the targeted gene-specific primers used to assess potential knockdown , primers for S . mansoni glyceraldehyde 3-phosphate dehydrogenase ( GAPDH ) and α-tubulin were used as endogenous normalization controls in all samples tested . Other controls for verifying the specificity of RNA treatment effects included ( 1 ) larval treatment with irrelevant GFP dsRNA and ( 2 ) treatment with a nontarget S . mansoni dsRNA . Finally , each q-PCR run was performed with 2 internal controls assessing both potential genomic DNA contaminations ( no reverse transcriptase added ) and purity of the reagents used ( no cDNA added ) . For each specific set of primers , all individual treatments ( including specificity controls ) were run in three technical replicates . Each experiment was repeated 3–5 times ( N = 3–5 ) as independent biological replicates and the ΔΔCt method [36] , using GAPDH and α-tubulin as endogenous loading controls to normalize the quantification of all cDNA targets was used to quantitatively compare treatment and control steady-state transcript levels . Due to the nonparametric distribution of data , statistical analysis of ΔΔCt values was performed using the Mann-Whitney U-test with significance set at P≤0 . 05 . Using a sporocyst-reactive rabbit anti-elongation factor 1α ( anti-EF1α ) antibody ( Cell Signaling Technology , Danvers , MA ) , Western blot analysis and immunocytochemical localization experiments were performed to monitor EF1α protein levels in EF1α dsRNA-treated sporocysts . For Western blots miracidia , transformed in the presence of 50 nM EF1α or control GFP dsRNA and cultivated in vitro for 7 days , were extracted in Trizol reagent ( Invitrogen ) as previously described , separated by standard 12 . 5% SDS-PAGE methods [37] , and electroblotted to nitrocellulose membranes ( Biorad Lab , Richmond , CA ) using a semi-dry protein transfer apparatus ( Hoefer TE 70 , Amersham Biosciences ) . After transfer , membranes were blocked overnight in TBS ( 2 . 42 g Tris base and 8 g NaCl/L , pH 7 . 6 ) containing 5% bovine serum albumin ( BSA ) , followed by incubation in a mixture of anti-EF1α ( 1∶ 1000 ) and rabbit anti-SmGST26 ( loading control , 1∶1000 dilution; Cell Signaling Technology ) , for 16 hr at 4°C . Membranes were washed 3 times in TBS-Tween ( 0 . 1% ) and incubated for 1 hr in TBS 5% BSA containing either goat anti-rabbit IgG-alkaline phosphatase-tagged ( AP ) or AP-labeled goat anti-rabbit IgG ( 1∶104 and 1∶5000 , respectively ) . Colorimetric immunoreactivity was detected with the chromogen substrate 5-bromo-4-chloro-3-indolyl phosphate ( BCIP ) and nitro-blue tetrazolium ( NBT ) , diluted in AP buffer ( 0 . 1 M Tris , 0 . 1 M NaCl , 0 . 05 M MgCl2 , pH 9 . 5 ) . To quantify the relative levels of anti-EF1α in specific dsRNA- and control GFP dsRNA-treated sporocyst extracts , target and control immunoreactivities were measured using an Ultraviolet Trans-illuminator BioImaging Systems ( UVP , Inc . , Upland , CA ) with the co-processed anti-GST26 band serving as a normalizing signal ( loading control ) . Quantitative comparisons of protein expression for EF1α in control and target dsRNA-treated sporocysts were analyzed by LabWorks Image Acquisition and Analysis Software version 4 . 6 . For immunocytochemical studies , 7-day old dsRNA-EF1α or -GFP-treated sporocysts were washed in CBSS ( allowing removal of ciliated epidermal plates ) , transferred to siliconized-microcentrifuge tubes in 500 µL PT buffer ( 2% paraformaldehyde , 1% Triton-X100/sPBS ) , and incubated overnight at 4°C with constant rotation . Fixed-parasites were washed 5 times in sPBS by centrifugation at 1600 rpm ( 2 min ) , resuspended in 500 µL of blocking buffer ( 5% normal goat serum/0 . 02% azide/sPBS ) for 16 hr , under constant agitation before addition of rabbit-anti-EF1α primary antibodies ( 1∶200 dilution in blocking buffer ) and incubation overnight at 4°C . Parasites were washed for 10 min in sPBS , resuspended in 500 µL of AlexaFluor 488-conjugated goat anti-rabbit IgG ( 4 µg/mL blocking buffer ) and incubated for 16 hr at 4°C with agitation . Following antibody treatments , sporocysts were washed 5 times in sPBS by centrifugation ( 1600 rpm , 2 min ) , resuspended in 40 µL of sPBS and mounted on coverslips . Specimens were examined and photographed using a Nikon Eclipse TE2000 ( Nikon Instrument Inc . ) inverted microscope equipped with a Bio-Rad Radiance 2100 MP Rainbow Confocal/Multiphoton Imaging System ( W . M . Keck for Biological Imaging , Instrumentation , UW-Medical School ) .
In our initial phenotype analysis of 24 different dsRNA treatments , no differences between specific dsRNA-treated larvae and controls ( GFP dsRNA-treated and untreated sporocysts ) were noted in miracidial transformation rates , larval motility or mortality during the in vitro cultivation period . The only notable phenotype observed in treated 7-day cultured sporocysts was an apparent greater number of small-sized or shortened sporocysts possibly involving a growth-related defect ( s ) ( Fig . 1A , 1B ) . However , because sporocysts in a given culture population typically represented a range of sizes , live sporocyst images were captured , from which larval lengths were measured and digitally-analyzed using Metamorph software . Within each biological replicate , such measurements were taken for the dsRNA treatment groups and statistically compared to both the GFP dsRNA-treated and no treatment ( blank ) groups . For a given dsRNA to be identified as having a putative dsRNA-mediated effect , the median larval length had to significantly differ ( P≤0 . 05 ) from both the GFP dsRNA and the blank controls in each of the biological replicates . Using these criteria , we observed significant decreases in parasite lengths in 9 of 24 dsRNAs in the first screening trial: Smad4 , lactate dehydrogenase ( LDH ) , Smad2 , Cav2A , elongation factor 1α ( EF1α ) , Smad1 , RHO2 , calcineurin B , and ring box ( Fig . 2 ) . Similar results were found in a second experimental series , which included nine additional dsRNA treatments . In this case , using the same criteria for significance , 2 of the 9 dsRNAs treatments ( GST26 and SOD ) exhibited a consistent size-related phenotype effect when compared to controls ( Fig . 3 ) . As before , greater frequency of shortened larvae was the only observable dsRNA-associated phenotype . To illustrate the variability in parasite response to the different dsRNA at the population level , and to underscore the importance of biological replication in phenotypic analysis of dsRNA treatment effects , 9 of the 33 dsRNA treatments ( Scmeg , HEXPB , Sm zinc finger1 ( SmZF1 ) , -calpain , -myosin light chain , PKCß , SPO1 , TPx1/2 and calcium channel ) exhibited significant length decreases in 2 of 3 experiments suggesting a possible , but inconsistent , connection with the observed phenotype . Sporocyst treatment by the remaining dsRNAs ( 14-3-3 protein , Sma2 , RHO1 , PKC receptor , DHHC , PEPCK , calreticulin , glycoprotein K5 , fibrillarin , GST28 , GPx , TPx1 , and TPx2 ) had no measurable effect on larval phenotype ( Figs . 2 and 3 ) . In order to document potential difference in dsRNA uptake within larval populations and between treatments , miracidia were exposed to rhodamine ( Rh ) -labeled dsRNA-GFP , -Smad4 or –LDH . After 7 days of exposure , dsRNA-uptake in sporocysts was assessed by fluorescent microscopy . As shown in Figure 4 , larvae within a single population exhibited wide variation in their abilities to take up labeled dsRNA , regardless of transcript species ( Figs . 4A and B ) . Indeed , a one-way ANOVA comparing the 3 dsRNA-treated groups for the prevalence of tegumental or internal staining was non-significant ( F = 1 . 159; P = 0 . 2555 ) indicating no differences in staining distribution between groups or locations . The most prominent sites of Rh-dsRNA localization in positively-stained sporocysts ( 67% of larvae ) were in the tegument ( ∼28% ) , excretory pores/flame cells and in unidentified parenchymal-like cells ( ∼39% ) ( Fig . 4C ) . Negative controls consisting of larvae treated with unlabeled-dsRNA did not display any fluorescent signal ( data not shown ) . Sporocysts within a given population exhibited heterogeneous ( +/− ) Rh-staining indicating specific dsRNA uptake by larvae , and not a nonspecific uptake via Rh-binding . Because the phenotypic screen revealed both phenotype-associated and nonphenotype-associated dsRNA species , we selected a subset of 24 genes to assess the effect of dsRNA treatments on steady-state transcript levels using real-time quantitative PCR ( q-PCR ) . Comparisons of normalized-levels of dsRNA-targeted messenger RNAs to their corresponding control treatment ( GFP dsRNA-treated group ) resulted in 12 transcripts that exhibited significantly reduced expression levels ( Fig . 5 ) . Unexpectedly , SOD transcripts consistently increased , rather than decreased , to very high levels of expression ( >1600-fold ) upon specific dsRNA treatment . A comparison was made between dsRNA species that produced a detectable phenotype and those generating a significant transcript knockdown ( or induced expression ) in an attempt to directly correlate phenotype and gene expression . Notably , only 7 of the 11 target dsRNA-treatments that produced a “shortened” larval phenotype presented a significant alteration in transcript levels ( Smad4 , lactate dehydrogenase , Cav2A , EF1α , calcineurin B , GST26 and SOD ) when compared to dsRNA-GFP treated sporocysts ( Fig . 5 ) . Smad4 and LDH dsRNA treatments showed a small , but consistently significant 15% decrease , while Cav2A , calcineurin B , EF1α and GST26 exhibited knockdowns of 42% , 65% , 70% and 85% , respectively . SOD , whose transcript levels were dramatically increased in dsRNA-treated sporocysts , also was phenotype-associated . In addition , as noted in Figure 5 , 6 dsRNA treatment groups that did not exhibit significant or consistent changes in larval length expressed significantly lower transcript levels than controls ranging from an approximately 30% ( SmZF1 ) to 75% ( fibrillarin , GST28 , GPx , TPx1 , and TPx2 ) after 7 days of exposure . No changes in transcript levels were observed for phenotype-associated Smad1 , Smad2 , RHO2 and ring box dsRNA treatments and phenotype-non-associated myosin , PKCB , Pepck , calpain , 14 . 3 . 3 protein , glycoprotein K5 , and SPO1 dsRNAs . Since we typically used day 7 as our temporal end-point for assessing RNAi phenotypic effect , we also investigated the possibility that transcripts may have been knocked down prior to day 7 . Using a subsampling of dsRNA species that represented a range of transcript knockdown levels , S . mansoni miracidia were treated with dsRNA-EF1α , -calcineurin B , -SOD , -LDH , -RHO2 , -Smad2 -Smad4 , myosin light chain and -ring box , and sporocyst transcript levels analyzed after 2 and 4 days postexposure . Compared to our previous 7-day treatment effects , results yielded various patterns of transcript silencing ( Fig . 6 ) . For example , although EF1α and calcineurin B transcripts were significantly reduced by ∼70% by day 7 , calcineurin B knockdown was actually greatest ( ∼90% ) at 2 days postexposure to dsRNA . Smad4 and LDH mRNAs , which previously showed a small , but significant , decrease at day 7 exhibited highest knockdown ( ∼40% ) on day 2 indicating an early effect of dsRNA treatment . SOD was found to be over-expressed at all time points , with an initial increase of ∼1200% at day 2 , followed by a maximum ∼17 , 000-fold expression at day 4 , before again returning to day-2 levels after 7 days of exposure . SOD transcript levels , however , were unaffected by heterologous exposure of larvae to several non-SOD-related dsRNAs ( data not shown ) . RHO2 dsRNA , previously displaying no effect on homologous transcript expression in sporocysts at 7 days , showed significant transcript knockdown at 2 and 4 days post-exposure before recovering to control levels at the 7-day time point . In contrast , Smad2 , myosin light chain and ring box dsRNA treatments demonstrated no consistent effect on their respective transcript levels regardless of the sampling interval . Finally , because we had available an antibody that was specifically reactive to the S . mansoni EF1α protein , we assessed the effect of EF1α dsRNA treatment on EF1α protein levels using Western blot and immunofluorescence imaging . Western blot analysis clearly showed that EF1α dsRNA-treated 7-day sporocyst extracts were significantly reduced in EF1α protein ( 50 kDa band ) compared to the GFP dsRNA-treated control group ( Fig . 7 ) . The presence of a 25 kDa GST26 band ( used as an antibody specificity and loading control ) in both the EF1α and GFP dsRNA-treated samples suggested both a specific EF1α transcript silencing and associated protein knockout ( Fig . 7 ) . This result was quantitatively confirmed by densitometry showing that , following normalization of transcripts to the loading control , EF1α protein was highly reduced by >80% in the EF1α dsRNA-treated sample compared to the GFP dsRNA control . Confocal immunolocalization of EF1α in intact dsRNA-treated sporocysts was consistent with the Western blot analysis: EF1α dsRNA-treated larvae displayed little immunoreactivity , while abundant anti-EF1α-reactivity was evident within cells and parenchymal tissues of GFP dsRNA-treated sporocysts ( Fig . 8 ) .
RNA interference ( RNAi ) has been widely used in a variety of organisms as a reverse-genetic approach to generate functional gene knockdowns with associated phenotypic changes [38]–[40] . In combination with complete and well-annotated genome databases , tools developed for RNAi now permit systematic , whole-genome screening leading to putative functional assignments for unknown genes , or direct functional confirmation of genes identified by sequence homology ( orthologues ) [41] . Some RNAi libraries are already available and have taken advantage of this functional genomics approach including for D . melanogaster [39] and C . elegans [42] , [43] . However , many newly-defined functions for any given gene tend to be organism-specific and may not always be identical to , or even homologous with , a similar gene's function in other species . Therefore , one of the current challenges we face is to integrate this organism-specific RNAi-derived functional information into the existing , ever-growing genomic databases of diverse organisms [44] . In addition , as noted by Geldhof [31] , application of RNAi approaches to parasitic helminths have at times lacked convincing evidence of an RNAi effect or have not provided information on the full spectrum or diversity of target transcript susceptibilities to dsRNA treatments . The RNAi screening approach described in the current study , to our knowledge , is the first to profile morphological phenotypes associated with exposure of larval schistosome blood flukes to dsRNAs representing a diversity of expressed genes . From our sampling of dsRNAs for 32 different S . mansoni genes known to be expressed in primary sporocysts , only 34% ( 11 transcripts ) produced a consistent , highly reproducible phenotype; namely a reduced larval length ( shortening ) , morphologically resembling a type of growth inhibition . Interestingly , this was the same phenotype that was observed in an earlier study involving dsRNA-mediated knockdown of a CD36-like scavenger receptor at the tegumental surface of S . mansoni sporocysts [27] . Thus , the genes associated with this phenotype are quite varied , including signaling molecules , Ca-interactive proteins , redox enzymes and a membrane receptors/ion channels . Although it would be premature to speculate on specific gene-phenotype linkages , it may not be particularly surprising that such a general phenotype as larval size might be regulated by many different genes expressed in variety of cell or tissue-types . In whole genome RNAi studies of C . elegans , an overall ∼2% of detectable viable phenotypes were growth- or size-related [41] and recent RNAi applications on schistosomules , miracidia , or adults also produced a similar consistent “shorter” phenotypes [45] . Our finding that only a proportion of sporocyst exhibited the shortened phenotype might be explained , at least in part , by results of the rhodamine-labeled dsRNA uptake experiment demonstrating that ∼67% of dsRNA-treated larvae within a population ( in all treatments ) exhibited signs of labeling , and of those , cellular localization of Rh-dsRNA within sporocysts varied considerably ( tegument , flame cells , parenchymal tissues ) . Thus the degree and site of dsRNA penetrance may be among several critical determinants influencing the observed phenotype . Moreover , this differential dsRNA uptake also could explain the variation in the levels of transcript knockdown observed in q-PCR analyses . Attempts to correlate phenotype and knockdown of target gene expression also yielded variable results in that 7 of the 11 genes associated with the shortened phenotype were significantly altered in their expression after the treatment period . One explanation as to why all phenotype-expressing transcripts were not reduced is the possibility that some genes possess different kinetic profiles ( i . e . , may have exhibited knockdown prior to day 7 ) . Of 5 genes whose transcript levels were unaffected ( Smad2 , Rho2 , ring box ) or marginally affected ( Smad4 , LDH ) by dsRNA at 7-days of treatment , 3 transcripts ( Rho2 , Smad4 and LDH ) showed an early significant knockdown at day 2 suggesting a temporal reduction in transcript levels that could be phenotype-associated . Why transcript expression of the other 2 phenotype-associated genes ( Smad2 and ring box ) was not affected by specific dsRNA treatments remains unclear , although RNAi off-targeting , i . e . , a mis-targeting of specific dsRNA to other unidentified mRNAs [46]–[49] could be involved . Off-target effect of introduced dsRNAs seems to be a common occurrence in helminth RNAi experiments , and presents a challenge in controlling such effects , as recently reviewed by Geldhof and colleagues [31] . Yet , it has been shown in mammal cells that knockdown efficiency is highly dependent on the specific dsRNA sequence of a particular gene [50] , and that in some cases , a small degree of similarity may invoke off-target gene silencing [51] . In our current study , we exposed larvae to gene-specific long ( 500 bp ) dsRNA , which , upon dicer cleavage , results in short unpredictable RNA sequences that represent potential sources of off-target gene silencing . To complicate matters further , some siRNA also have been shown to exhibit nonspecific toxic effects that may directly affect transcriptional processes without altering specific transcript levels [52] . Although it was not the goal of this study to provide in-depth analyses of each the gene investigated herein , followup experiments involving S . mansoni elongation factor 1α ( EF1α ) illustrates the importance of providing several lines of evidence of an RNAi effect . In this case , larval treatment with EF1α dsRNA resulted in a demonstrable phenotype , specific transcript knockdown , and approximately 80% inhibition of EF1α protein expression as measured by both Western blot and immunocytochemical assays . Elongation factors are known to be essential in the translational process by functioning to catalyze the aminoacyl-tRNA delivery to ribosomes during protein elongation [53] . Given this putative function of EF1α , and its widespread knockdown at the transcript and protein levels , its involvement either directly or indirectly in generating the shortened larval phenotype is supported by the data presented here . Reasons why we did not see any changes in GST26 protein levels in the EF1α Western blot assay maybe due to a slow protein turnover rate for GST26 or the possibility that GST26 is synthesized in cells/tissues that were unaffected by EF1α dsRNA knockdown . Variation in dsRNA treatment effects also was evident in the group of sporocysts that lacked any definable phenotype . Of the 13 nonphenotype-expressing larval groups , half ( 6 ) exhibited consistent , significant reductions in transcript levels as measured by q-PCR , while transcript levels in the other half ( 7 ) were unaffected . This type of result is not unexpected as this has been demonstrated previously in RNAi screens of model organisms such as C . elegans [54] , as well as parasitic nematodes [31] , [55] . For those transcripts whose expression was unaffected by specific dsRNA treatment , there would not be an a priori expectation of phenotypic change . There are several ways to potentially explain a lack of differential phenotype in larvae presenting with dsRNA-induced transcript knockdown : ( 1 ) the gene targeted for dsRNA knockdown is functionally unrelated to the observed phenotype , ( 2 ) since typically an RNAi-like effect does not lead to a complete gene ( and presumably protein ) knockout , sufficient protein synthesis/activity remains to continue support of the normal “phenotype” , ( 3 ) other related proteins and/or isoforms may be supplementing or replacing the protein ( and its function ) initially targeted for dsRNA-mediated silencing , and ( 4 ) the protein product of the targeted transcript may have a lengthy half-life ( i . e . , slow turnover rate ) , hence delaying potential gene knockdown effects at the protein level . These results further underscore the wide variation in susceptibilities of individual S . mansoni genes to RNAi procedures , and the fact that dsRNA knockdown may not be associated with any demonstrable phenotype . One of the more intriguing results of our study was the consistent , high level upregulated expression of SOD in sporocysts upon treatment with SOD dsRNA . Even more impressive than the 1600-fold transcript expression following the standard 7-day incubation period was the ∼17 , 000-fold expression 2 days earlier ( day 4 ) . At present we do not have an explanation as to how larval exposure to SOD dsRNA may be triggering such high expression levels . One possibility is that the yet unknown sequence ( s ) in processed SOD dsRNA may be stimulating reactions similar to RNA activation ( RNAa ) [56] , [57] . If this is indeed the case , this would be a novel facet of RNAi in parasitic helminths . Although the function of SOD as a protective anti-oxidant has been suggested [58]–[60] , its essential role in parasite development has not been established . If the overexpression response seen in this study is linked to SOD depletion or SOD sequence activation , this would imply a critical role in sporocyst survival , and perhaps in larval development as evidenced by its association with the sporocyst size phenotype . The role of endogenous SOD , and other anti-oxidants , in sporocysts confronted with oxidative stress is the subject of ongoing investigations in our laboratory . To date , RNAi is the only reverse genetic tool available in schistosomes [22] , and although it has been successfully applied as a functional genomics tool in both mammalian [25] , [28] , [61] and snail [26] , [27] stages of infection , a lack of consistency in the RNAi-induced knockdown and resulting phenotypes indicates a pressing need to more fully investigate RNAi to gain a better understanding of this complex mechanism in S . mansoni , and other parasitic flatworm species [31] . The results presented here provide an overview of the variability that may be encountered as transcript-specific dsRNA sequences are applied as a tool for targeted gene manipulation and morphological phenotyping in larval schistosomes . It is anticipated that further improvements in dsRNA delivery methods likely would be beneficial in attaining more consistent transcript/protein knockdowns and resulting phenotypes , as will further detailed analyses of specific siRNA for individual genes . Future studies involving optimization of transfection reagent- and electroporation-based gene delivery approaches are currently being planned . In assessing RNAi effects , in addition to low and/or inconsistent dsRNA penetrance , we also are hindered by the numbers of parasites that can be processed for dsRNA treatments at a given time , and , as reported in this study , a very limited phenotype repertoire , due to a lack of more sensitive detection tools . These restrictions illustrate some of the limitations facing large-scale RNAi experiments , and demonstrate the necessity of small-scale or gene-by-gene characterizations , until development of more sensitive , higher-throughput methodologies [44] . In summary , this study is the first to provide a multi-gene assessment of the efficacy of dsRNA treatments in characterizing phenotypic and transcriptional changes brought about by introduction of gene-specific dsRNAs into cultured S . mansoni larvae . Prolonged exposure to dsRNA , when selectively applied to target genes expressed in early larval stages , can generate significant transcript knockdown , thus facilitating the investigation of potential gene-associated function . However , as shown in the present study , individual genes may differ significantly in their abilities to render RNAi-like effects , and this is likely due , at least partially , to efficacy of their intracellular processing . Although RNAi approaches continue to be potentially valuable tools for functional genomics in parasitic helminths , caution should be taken in the design , set-up and execution of RNAi experiments . As a followup to this study , we are now focusing on the group of enzymes involved in reduction-oxidation ( redox ) reactions , especially those with antioxidant activity , and that have exhibited consistent transcriptional knockdown by RNAi . These functional studies were made possible by the data provided in this initial multi-gene profiling of dsRNA effects . | RNA interference ( RNAi ) represents the only method currently available for manipulating gene-specific expression in human blood flukes , Schistosoma spp . , although its application as a functional genomics tool in early intramolluscan larval stages has been limited to single gene analyses . Accelerating gene discovery efforts over the past 10 years have resulted in extensive , ever-increasing databases of genomic , transcriptomic and EST sequences . Unfortunately , our understanding of the function of the vast majority of these genes has not kept pace with their discovery , and this represents a significant barrier and the next real challenge for investigators of schistosomes , and other parasitic helminths . In the present study , we selected an array of 32 genes expressed in S . mansoni sporocysts to evaluate their susceptibility to double-stranded ( ds ) RNA treatment and to begin characterizing morphological phenotypes associated with a potential RNAi effect . Results demonstrate that gene knockdown and/or resulting phenotypes are highly transcript-dependent ( specific dsRNA sequence used ) and vary with time post-dsRNA exposure . Because of this potential variability in both transcript and phenotype expression in response to dsRNA treatment , our findings illustrate that , although a RNAi-type approach holds great promise as a functional reverse-genetics tool for larval schistosomes , its application requires caution in the design and execution of experiments and interpretation of results . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [
"molecular",
"biology",
"cell",
"biology/developmental",
"molecular",
"mechanisms",
"cell",
"biology/gene",
"expression",
"genetics",
"and",
"genomics/functional",
"genomics"
] | 2009 | Phenotypic Screen of Early-Developing Larvae of the Blood Fluke, Schistosoma mansoni, using RNA Interference |
Cancer development and progression result from somatic evolution by an accumulation of genomic alterations . The effects of those alterations on the fitness of somatic cells lead to evolutionary adaptations such as increased cell proliferation , angiogenesis , and altered anticancer drug responses . However , there are few general mathematical models to quantitatively examine how perturbations of a single gene shape subsequent evolution of the cancer genome . In this study , we proposed the gene gravity model to study the evolution of cancer genomes by incorporating the genome-wide transcription and somatic mutation profiles of ~3 , 000 tumors across 9 cancer types from The Cancer Genome Atlas into a broad gene network . We found that somatic mutations of a cancer driver gene may drive cancer genome evolution by inducing mutations in other genes . This functional consequence is often generated by the combined effect of genetic and epigenetic ( e . g . , chromatin regulation ) alterations . By quantifying cancer genome evolution using the gene gravity model , we identified six putative cancer genes ( AHNAK , COL11A1 , DDX3X , FAT4 , STAG2 , and SYNE1 ) . The tumor genomes harboring the nonsynonymous somatic mutations in these genes had a higher mutation density at the genome level compared to the wild-type groups . Furthermore , we provided statistical evidence that hypermutation of cancer driver genes on inactive X chromosomes is a general feature in female cancer genomes . In summary , this study sheds light on the functional consequences and evolutionary characteristics of somatic mutations during tumorigenesis by propelling adaptive cancer genome evolution , which would provide new perspectives for cancer research and therapeutics .
Cancer development and progression are mediated by the accumulation of genomic alterations , including point mutations , insertions and deletions , gene fusions , amplifications , and chromosomal rearrangements [1 , 2] . The majority of the somatic mutations found in tumor cells are ‘passenger’ rather than ‘driver’ mutations [3] . In 1976 , Peter Nowell wrote a landmark perspective for the clonal evolution model of cancer and applied evolutionary models to understand tumor growth and treatment failure [4] . He proposed that most neoplasms arise from a single cell , and tumor progression results from acquired genetic variability within the original clone , allowing sequential selection of more aggressive sublines . He also noted that genetic instability , occurring in tumor cells during disease progression , might enhance this process . This view now has been widely accepted [4 , 5] . Somatic cell evolution leads to adaptive cancer cell survival , including increased proliferative , angiogenic , and invasive phenotypes [2] . However , understanding how somatic cell evolution drives tumorigenesis remains a great challenge in cancer research . Genome instabilities , such as chromosomal instability and microsatellite instability , have been well studied in cellular systems [2 , 6 , 7] . For example , Teng et al . found that in yeast a mutation on a single gene may cause genomic instability , leading to adaptive genetic changes [8] . Whether and how human tumor genomes are genetically unstable , induced by single gene alterations , has been debated for decades [9–12] , but has recently gained much support . For instance , Emerling et al . found an amplification of PIP4K2B in HER-2/Neu-positive breast cancer with its co-occurrence with mutations in TP53 [11] . They showed that a subset of breast cancer patients had a high level of gene expression of PIP4K2A and PIP4K2B and provided evidence that these kinases are essential for growth in the absence of p53 . Liu et al . found that POLR2A ( encoding the largest and catalytic subunit of the RNA polymerase II complex ) was deleted together with TP53 in cancer cell lines and primary tumors in human colon cancer [13] . Additionally , the DNA cytidine deaminase APOBEC3B-catalyzed genomic uracil lesions are responsible for a large proportion of both dispersed and clustered mutations in multiple distinct cancers [12] . These lines of evidence show that single gene alterations may induce the mutations of other genes in a cancer genome that drive tumorigenesis and tumor progression [9–13] . Thus , a quantitative assessment of whether the perturbation of any single gene in a cancer genome is sufficient to drive genetic changes would help us better understand tumorigenesis and tumor evolution through genomic alterations . However , distinguishing functional somatic mutations from massive passenger mutations and non-genetic events is a major challenge in cancer research . Massive genomic alterations present researchers with a dilemma: does this somatic genome evolution contribute to cancer , or is it simply a byproduct of cellular processes gone awry [14] ? Cells consist of various molecular structures that form complex , dynamic , and plastic networks [15] . In the molecular network framework , a genetic aberration may cause network architectural changes through affecting or removing a node or its connection within the network , or changing the biochemical properties of a node ( protein ) [16–18] . The abundance of next-generation sequencing data of cancer genomes provides biologists with an unprecedented opportunity to gain a network-level understanding of tumorigenesis and tumor progression [15 , 19–22] . However , how to integrate large-scale molecular networks with cancer genomic aberrations is highly challenging [9 , 10] . The development of a mathematical model will be helpful to understand how genetic aberrations perturb the molecular network architecture and manifest the effects during tumorigenesis . In this study , we proposed a novel mathematical model , namely gene gravity model , derived from Newton’s law of gravitation to study the evolution of cancer genomes . The gene gravity model detects a gene-gene pair that two genes are co-mutated and highly co-expressed simultaneously in a given cancer type based on several previous evidences [8 , 11 , 13] . As proof of principle , we applied the model to approximately 3 , 000 tumors’ transcription and somatic mutation profiles across 9 cancer types from The Cancer Genome Atlas ( TCGA ) project . We found that cancer driver genes may shape somatic genome evolution by inducing mutations in other genes during tumorigenesis . We identified six putative cancer genes by quantifying the gene gravity model . Furthermore , we found a higher somatic mutation density related to cancer driver genes on the X chromosome in comparison to the whole autosomes , suggesting that hypermutation in inactive X chromosomes is a general feature in females . In summary , this study would provide new insights into adaptive cancer genome evolution shaped by somatic mutations in cancer .
The gene gravity model postulates that if two genes have high mutation density and strong gene co-expression in a given cancer type , they should have a higher G score and related to a higher risk of inducing mutations to other genes; this postulation is based on several previous observations [8 , 11 , 13] . We developed the gene gravity model by incorporating ~3 , 000 tumors’ transcription and somatic mutation profiles across 9 cancer types from TCGA under molecular network architecture knowledge ( Fig 1 ) . These 9 cancer types consist of breast invasive carcinoma ( BRCA ) , colon adenocarcinoma ( COAD ) , glioblastoma multiforme ( GBM ) , head and neck squamous cell carcinoma ( HNSC ) , kidney renal clear cell carcinoma ( KIRC ) , lung adenocarcinoma ( LUAD ) , lung squamous cell carcinoma ( LUSC ) , ovarian serous cystadenocarcinoma ( OV ) , and uterine corpus endometrial carcinoma ( UCEC ) . First , we collected 3 , 487 tumor transcription profiles ( RNA-Seq ) for the 9 cancer types . Then , we constructed 9 co-expressed protein interaction networks ( CePINs ) for the 9 cancer types ( S1 Table ) respectively by incorporating the transcription profiles into a large-scale protein interaction network ( PIN ) in S2 Table and Fig 1A . Each CePIN contained ~100 , 000 edges connecting ~12 , 000 genes . Second , we collected 277 , 370 nonsynonymous somatic mutations identified from 2 , 946 tumor exomes across 9 cancer types from TCGA ( S1 Table ) . For each cancer type , we projected the somatic mutations onto PIN to construct a somatic mutation PIN via a network propagation algorithm ( Fig 1B and 1C ) . We then derived a G score for each gene-gene pair in the 9 cancer types , using Newton’s law of gravitation ( Fig 1C ) . Then , we examined the G score for seven gene sets: cancer driver genes , cancer gene census ( CGC ) genes ( experimentally validated cancer genes ) , tumor suppressor genes ( TSGs ) , oncogenes , DNA repair genes , chromatin regulation factors ( CRFs ) , and essential genes ( Fig 1D ) . Finally , we investigated the pattern of hypermutation of the inactive X chromosome in female versus male cancer genomes by quantifying cancer genome evolution using the gene gravity model ( Fig 1E ) . To verify the gene gravity model , we investigated the enrichment of somatic mutations on protein-protein interaction ( PPI ) pairs as well as unfiltered interactions relative to the same number of random pairs based a previous study [23] . We found that PIN is significantly more enriched for high mutation density than random pairs across the 9 cancer types ( q < 2 . 2 × 10−16 , Wilcoxon rank-sum test corrected by Benjamini-Hochberg multiple testing , S1 Fig ) . We first examined the distribution of G score for two benchmark gene sets: DNA repair genes and CRFs . The CRFs modulating the epigenetic landscape have emerged as potential gatekeepers and signaling coordinators for the maintenance of genome integrity [24] . The enzymes encoded by DNA repair genes continuously monitor chromosomes to repair damaged nucleotide residues generated by exposure to carcinogens and cytotoxic agents ( e . g . , anticancer drugs ) [25] . Thus , both CRFs and DNA repair genes are of critical importance for the maintenance of the genetic information in the cancer genome . In this study , we collected two high-quality gene sets: 153 DNA repair genes [26] and 176 CRFs [27] ( S3 Table ) . We defined a DNA repair gene-gene pair gravitational interaction as one or two genes in a pair is/are DNA repair genes . A non-DNA repair gene-gene pair gravitational interaction was defined as neither of the two genes in a pair is a DNA repair gene . We applied the same definition for the remaining 6 gene sets: cancer driver genes , CGC genes , TSGs , oncogenes , CRFs , and essential genes . We then investigated the complementary cumulative G score ( S2–S10 Figs ) . We found that the DNA repair gene cumulative G score is higher than that of non-DNA repair genes in 8 cancer types , except BRCA . Furthermore , the CRF cumulative G score is higher than that of non-CRFs in all of the 9 cancer types ( S2–S10 Figs ) . Collectively , these observations demonstrated that we could use the gene gravity model to quantitatively examine how perturbations of a single gene shape subsequent evolution of the cancer genome based on evidence in several previous biological studies [8 , 11 , 13] . We investigated “high somatic evolutionary pressure” for a particular gene that tends to be co-mutated and highly co-expressed with other genes in a given cancer type . We hypothesized that if a gene has a higher somatic evolutionary pressure , this gene may increase subsequent genetic changes [8 , 11 , 13] . We compiled a high-quality , mutated cancer driver gene set ( 614 cancer driver genes , S3 Table ) from four pan-cancer genomic analysis projects [3 , 28–30] . We found that the cancer driver gene cumulative G score is significantly higher than that of non-cancer driver genes in all of the 9 cancer types ( q < 2 . 2 × 10−16 , Wilcoxon rank-sum test , S2–S10 Figs ) . These observations suggest that cancer driver mutations may increase subsequent genetic changes based on the previous studies [8 , 11 , 13] . We also studied CGC genes , which are well curated and have been widely used as a reference cancer gene set in many cancer-related studies [31 , 32] . As expected , we found that the CGC gene cumulative G score is higher than that of non-CGC genes in 6 cancer types: BRCA , COAD , GBM , HNSC , KIRC , and UCEC ( S2–S6 and S10 Figs ) . However , the CGC gene cumulative G score is slightly higher than that of non-CGC genes in 3 cancer types: LUAD , LUSC , and OV ( S7–S9 Figs ) . A previous study indicated that an average mutation frequency in smokers is more than 10-fold higher in never-smokers in non-small cell lung cancer [33] . We next separated TCGA patients into smokers and never-smokers in LUAD and LUSC , and reexamined the CGC gene cumulative G score . As expected , the CGC gene cumulative G score is significantly higher than that of non-CGC genes in LUAD and LUSC never-smokers ( q < 0 . 05 , S11 Fig ) . However , the CGC gene cumulative G score is slightly higher than that of non-CGC genes in LUAD and LUSC smokers ( S11 Fig ) . Thus , heterogeneous mutation frequencies and gene transcription profiles in the combined smokers and never-smokers in LUAD or LUSC may influence the performance of the gene gravity model [33] . For OV ( S9 Fig ) , high genomic instability of the ovarian cancer genome may cause this slight gene cumulative G score between CGC and non-CGC genes [34] . Finally , we considered essential genes . We compiled 2 , 719 essential genes ( S3 Table ) from the Online GEne Essentiality database [35] . S2–S10 Figs showed that the essential gene cumulative G score is higher than that of non-essential genes across 9 cancer types . Remarkably , the cancer driver gene-gene G score is higher than that of essential genes ( q < 0 . 01 ) in all of the 9 cancer types ( S2–S10 Figs ) . Tumorigenesis is dependent on the accumulation of one or multiple driver mutations that activate oncogenic pathways or inactivate tumor suppressors [36 , 37] . Oncogenes often positively co-expressed with interacting partners due to gain-of-function mutations; while TSGs often negatively co-expressed with interacting partners due to lose-of-function mutations [38] . Thus , we defined attractive gravitation ( AG ) as two genes that have positive gene co-expressed correlation and repulsive gravitation ( RG ) as two genes that have negative gene co-expressed correlation in a specific cancer type . We compiled 477 oncogenes and 1 , 040 TSGs ( S3 Table ) , and then examined the AG and RG score for oncogenes and TSGs , respectively . We found that the oncogene AG cumulative distribution is higher than that of non-oncogenes in 5 cancer types: BRCA , COAD , KIRC , OV , and UCEC ( S12 Fig ) . However , as shown in S13 Fig , the oncogene RG cumulative distribution is similar or slightly higher than that of non-oncogenes in all of the 9 cancer types . Additionally , we examined the AG and RG score for TSGs . We found that both AG and RG cumulative distribution for TSGs is higher than that of non-TSGs in 7 cancer types , except LUSC and OV ( S14 and S15 Figs ) . Taken together , our gene gravity model can distinguish one important tumor biological characteristics , oncogenic potential altered by oncogenes , very well . However , our model fails to distinguish caretaker or gatekeeper roles altered by TSGs . One possible reason is that some TSGs have both tumor suppressor and oncogenic activities in different cancer types or cell types . For example , p21 , encoded by CDKN1A , plays both tumor suppressor activities and paradoxical tumor-promoting activities in cancer [39] . In addition , it is partially because TSGs have truncated mutations that may scattered in the gene region . Thus , further study will be needed for systematic investigation of the AG and RG score for TSGs , which we hope will be prompted by the findings herein . We calculated the gene average gravitation ( aveG ) score using ( ρ ) i = ∑j Gij / n between gene i and gene j ( j belongs to the set of gene i’s interacting partners ( n ) in PIN ) . We found that the aveG score of cancer driver gene is significantly higher than that of DNA repair , CGC , and essential genes in all of the 9 cancer types ( Fig 2 and S4 Table ) . For BRCA , the cancer driver gene aveG score ( 0 . 47 ± 0 . 02 ) is significantly higher than that of DNA repair genes ( 0 . 30 ± 0 . 03 , q = 1 . 9 × 10−4 ) , CGC genes ( 0 . 35 ± 0 . 02 , q = 1 . 1 × 10−4 ) , and essential genes ( 0 . 26 ± 0 . 01 , q = 2 . 3 × 10−32 , S4 Table ) . However , the cancer driver gene aveG score is similar to that of CRFs ( 0 . 42 ± 0 . 04 , q = 1 . 0 ) in BRCA . Similar trends were observed in the remaining 8 cancer types ( S4 Table ) . Thus , chromatin regulation might play an important role in tumorigenesis . We further investigated whether genetic or epigenetic alterations have combinatorial effects that shape cancer genome evolution . Since CRFs represent the epigenetic landscape [27] , we divided cancer driver genes into two subgroups: CRF cancer driver genes and non-CRF cancer driver genes . We found cancer driver genes are significantly enriched in CRFs ( 38 out 176 CRFs versus 176 CRFs from 20 , 462 human protein-coding genes collected from National Center for Biotechnology Information [NCBI] database , p = 3 . 0 × 10−21 , Fisher’s exact test , Fig 3A ) . Furthermore , the CRF cancer driver gene aveG score is higher than that of non-driver CRFs across 9 cancer types ( q < 0 . 10 , Fig 3A and S5 Table ) . For KIRC , the CRF cancer driver gene aveG score ( 1 . 6 ± 0 . 48 ) is significantly higher than that of non-CRF cancer driver genes ( 0 . 76 ± 0 . 04 , q = 4 . 2 × 10−3 ) and non-driver CRFs ( 0 . 72 ± 0 . 09 , q = 3 . 3 × 10−3 , S5 Table ) , respectively . However , we did not find a significant aveG difference between non-CRF cancer driver genes and non-driver CRFs in any of the 9 cancer types ( q = 1 . 0 , Fig 3A and S5 Table ) . We next divided CGC genes into two subgroups: CRF CGC genes and non-CRF CGC genes . We found that CGC genes are significantly enriched in CRFs as well ( p = 1 . 2 × 10−15 , Fisher’s exact test , S16A Fig ) . As expected , we did not observe a significant aveG difference between non-CRF CGC genes and non-CGC CRFs in 7 cancer types ( q > 0 . 05 , S6 Table ) , with the exception of OV ( q = 0 . 04 ) and KIRC ( q = 0 . 04 ) . Put together , the cancer genome evolution might be shaped by the combinatorial synergy between cancer driver genes and CRFs . We next divided cancer driver genes into two subgroups: DNA repair cancer driver genes and non-DNA repair cancer driver genes . Fig 3B showed that DNA repair genes tend to be cancer driver genes as well ( 18 out 153 DNA repair genes versus 153 DNA repair genes from 20 , 462 human protein-coding genes collected from NCBI database , p = 1 . 1 × 10−6 ) . However , CRFs are more likely to be cancer driver genes than DNA repair genes ( p = 0 . 02 ) . The DNA repair cancer driver gene aveG score is similar to that of non-DNA repair cancer driver genes in 6 cancer types ( q > 0 . 1 ) , except of HNSC ( q = 0 . 02 , S7 Table ) , KIRC ( q = 0 . 08 ) , and LUAD ( q = 0 . 08 ) . However , the DNA repair cancer driver gene aveG score is significantly higher than that of non-driver DNA repair genes ( q < 0 . 01 , S7 Table ) in all of the 9 cancer types ( Fig 3B ) . For BRCA , the DNA repair cancer driver gene aveG score ( 0 . 73 ± 0 . 13 ) is marginally higher than that of non-DNA repair cancer driver genes ( 0 . 46 ± 0 . 02 , q = 0 . 12 ) , while significantly higher than that of non-driver DNA repair genes ( 0 . 24 ± 0 . 03 , q = 4 . 4 × 10−4 , S7 Table ) . Furthermore , the non-DNA repair cancer driver gene aveG score is significantly higher than that of non-driver DNA repair genes in all of the 9 cancer types as well ( q < 0 . 01 , Fig 3B and S7 Table ) . We further divided CGC genes into two subgroups: DNA repair CGC genes and non-DNA repair CGC genes . We found CGC genes are significantly enriched in DNA repair genes as well ( p = 2 . 7 × 10−18 , Fisher’s exact test , S16B Fig ) . S8 Table indicated that DNA repair CGC gene aveG score is not significantly higher than that in both non-DNA repair CGC genes ( q > 0 . 50 ) and non-CGC DNA repair genes ( q > 0 . 10 ) in 8 cancer types with an exception of OV ( q = 0 . 03 ) . Moreover , the non-DNA repair CGC gene aveG score is higher than that of non-CGC DNA repair genes in COAD ( q = 0 . 04 ) and OV ( q = 0 . 02 , S8 Table ) . Collectively , the cancer genome evolution shaped by cancer driver genes may have additional mechanisms ( i . e . , chromatin regulation ) , except DNA repair . We found that the top 100 genes with the highest aveG scores tend to be cancer driver genes ( q < 0 . 01 , Fisher’s exact test , Fig 4A and S9 Table ) or CGC genes ( q < 0 . 05 , S10 Table ) in all of the 9 cancer types . In addition , the top 100 genes with the highest aveG scores are more likely to be CRFs ( q < 0 . 05 , S11 Table ) in 7 cancer types with the exception of COAD ( q = 0 . 12 ) and LUSC ( q = 0 . 12 ) . However , the top 100 genes are not significantly enriched in DNA repair genes in all of the 9 cancer types ( q > 0 . 05 , Fig 4A and S12 Table ) . We further examined the tumor exome mutation density ( the average number of mutations per Mb ) for the top 10 genes with the highest aveG score via the genome-wide mutation rate analysis ( S13 Table ) . By examining mutation density data of ~3 , 000 tumor exomes from Kandoth et al . [29] , we found that patients having nonsynonymous somatic mutations on any of four genes ( FAT4 , SYNE1 , AHNAK , or COL11A1 ) often showed a higher cancer genome mutation density at the whole genome level compared to that of wild-type ( WT ) patients in 4 cancer types: COAD , LUAD , LUSC , and UCEC ( Fig 4B ) . FAT4 ( protocadherin fat 4 ) , a member of the cadherin super-family , is a key component in the Hippo signaling pathway , playing a candidate tumor suppressor role in cancer [40] . In COAD , 40 patients harbored FAT4 nonsynonymous mutations . The average number of mutations per Mb for 40 FAT4 mutated COAD samples ( 43 . 3 ± 12 . 8 ) are significantly higher than that of FAT4 WT samples ( 5 . 0 ± 0 . 57 , q = 1 . 1 × 10−5 , Fig 4B ) . Similarly , the average number of mutations per Mb for 43 FAT4 mutated LUAD samples ( 26 . 3 ± 8 . 4 ) are significantly higher than that of FAT4 WT samples ( 7 . 5 ± 0 . 48 , q = 2 . 8 × 10−9 , Fig 4B ) . Using genome-wide association studies , Berndt et al . found FAT4 to be a candidate gene for spontaneous pulmonary adenomas [41] . Using exome sequencing , Zang et al . found that the somatic inactivation of FAT4 might be a critical tumorigenic event in a subset of gastric cancers [42] . In this study , FAT4 was identified as a putative cancer gene involved in lung and colorectal cancer , which is consistent with previous studies [40–43] . SYNE1 , encoding spectrin repeat containing , nuclear envelope 1 , is involved in nuclear organization and structural integrity , function of the Golgi apparatus , and cytokinesis . Herein , we found that the average number of mutations per Mb for 49 SYNE1 mutated COAD samples ( 35 . 8 ± 8 . 4 ) are significantly higher than that of SYNE1 WT samples ( 7 . 5 ± 0 . 48 , q = 6 . 8 × 10−9 , Fig 5B ) . Doherty et al . found that SYNE1 polymorphism relates to an increased risk of invasive ovarian cancer [44] . Collectively , SYNE1 may be a candidate cancer mutated gene in COAD . AHNAK ( neuroblast differentiation-associated protein ) , also known as desmoyokin , is essential for tumor cell migration and invasion [45] . In this study , the average number of mutations per Mb ( 12 . 1 ± 2 . 6 ) for 22 AHNAK mutated samples is significantly higher than that of AHNAK WT samples in HNSC ( 4 . 5 ± 0 . 21 , q = 1 . 5 × 10−5 , Fig 4B ) . Dumitru et al . found that AHNAK was associated with poor survival rates in laryngeal carcinoma , a major subtype of head and neck cancer [46] . COL11A1 and COL6A3 , encoding collagen proteins , are two main structural proteins of the various connective tissues in animals . In LUAD , the average number of mutations per Mb ( 25 . 3 ± 7 . 9 ) for 46 COL11A1 mutated samples is significantly higher than that of COL11A1 WT samples ( 7 . 4 ± 0 . 47 , q = 1 . 1 × 10−9 , Fig 5B ) . Additionally , for LUSC , the average number of mutations per Mb ( 16 . 5 ± 0 . 59 ) for 32 COL11A1 mutated samples is significantly higher than that of COL11A1 WT samples as well ( 8 . 5 ± 0 . 40 , q = 4 . 9 × 10−5 ) . Furthermore , COL6A3 ( q = 3 . 1 × 10−4 , COAD ) and COL5A2 ( q = 1 . 5 × 10−4 , LUAD ) mutations are significantly associated with a high mutation density in colorectal and lung cancer , respectively . The over-expression of COL11A1 reportedly correlates with lymph node metastasis and poor prognosis in non-small cell lung cancer and ovarian cancer [47–49] . The expression level of COL6A3 is involved in pancreatic malignancy [50 , 51] . Collectively , AHNAK , COL11A1 , and COL6A3 may be potential candidates for therapeutic and diagnostic biomarkers in head and neck cancer and lung carcinoma . However , the mutation status of each of aforementioned genes is associated with the genome-wide mutation rate . Mutations in these genes could be either the cause of the mutation-rate increase or simply a consequence of an elevated global mutation rate . Thus , further experimental validation of these genes in the specific cancer type is warranted . When examining cancer driver gene aveG score across chromosomes in each of 9 cancer types , interestingly , we found that the X chromosome has an unusually higher cancer driver gene aveG scores compared to autosomes in BRCA , GBM , and UCEC using the total 22 autosomes as background ( Fig 5 ) . In BRCA , cancer driver gene aveG score ( 0 . 66 ± 0 . 09 ) on the X chromosome is higher than that of the whole set of 22 autosomes ( 0 . 46 ± 0 . 02 , q = 0 . 06 [p = 7 . 9 × 10−3] , Wilcoxon rank-sum test , Fig 5A ) . Similarly , in GBM , the cancer driver gene aveG score ( 1 . 2 ± 0 . 18 ) on the X chromosome is higher than that of the whole set of 22 autosomes ( 0 . 80 ± 0 . 05 , q = 0 . 07 [p = 9 . 9 × 10−3] , Fig 5B ) . And the cancer driver gene aveG score ( 0 . 92 ± 0 . 15 ) on the X chromosome is also higher than that of the whole set of 22 autosomes in UCEC ( 0 . 56 ± 0 . 03 , q = 0 . 04 [p = 5 . 3 × 10−3] , Fig 5C ) . As a control , we repeated the aforementioned analyses for all genes and essential genes , respectively . We did not find the higher aveG score on the X chromosome for all genes or essential genes in any of the 9 cancer types ( Fig 5 and S17 Fig ) . Thus , the high gene aveG score on the X chromosome is unique for cancer driver genes . The X chromosome is largely functionally haploid in both males and females . A recent study showed that hypermutation of the inactive X chromosome is a frequent event in cancer [52] . Both BRCA and UCEC ( Fig 5 ) are female-specific cancer , while GBM is not . To explore the hypermutation of inactive X chromosome in the female versus male cancer genomes , we separated GBM patients as males and females , and performed the same analysis . Interestingly , we found that the cancer driver gene aveG score ( 0 . 66 ± 0 . 13 ) on the X chromosome is significantly higher than that of the whole set of 22 autosomes ( 0 . 43 ± 0 . 04 , q = 0 . 04 , Fig 6A and 6C ) in the female GBM genomes . However , the cancer driver gene aveG score ( 0 . 68 ± 0 . 17 ) on the X chromosome is similar to that of the whole set of 22 autosomes ( 0 . 72 ± 0 . 07 , q = 0 . 68 , Fig 6B and 6C ) in the male GBM genomes . Furthermore , similar aveG scores for all genes ( q = 0 . 09 ) or essential genes ( q = 0 . 18 ) were observed between the X chromosome and the whole set of 22 autosomes in the female GBM genomes . In contrast , we found a lower aveG score on the X chromosome for all genes ( q = 4 . 4 × 10−9 , Fig 6C ) or essential genes ( q = 0 . 06 ) compared to that on the whole set of 22 autosomes in the male GBM genomes . We then examined the top 10 driver genes with the highest aveG scores on the X chromosome in BRCA , GBM , and UCEC . Two putative cancer drivers ( DDX3X and STAG2 ) stood out ( Fig 6D and 6E ) . We found that the patients harboring DDX3X or STAG2 nonsynonymous mutations have a higher genome mutation density in uterine cancer during the genome-wide mutation rate analysis ( Fig 6D ) . For instance , the average number of mutations per Mb for 15 DDX3X mutated uterine tumors is 144 . 1 ± 34 . 0 , 11-fold higher than that of DDX3X WT tumors ( 13 . 1 ± 2 . 8 , q = 2 . 5 × 10−5 ) . A previous study indicated that somatic mutations of DDX3X were associated with medulloblastoma [53] . Additionally , the average number of mutations per Mb for 26 STAG2 mutated uterine tumors ( 144 . 5 ± 26 . 0 ) is significantly higher than that for STAG2 WT samples ( 10 . 2 ± 2 . 2 , q = 1 . 9 × 10−10 ) . STAG2 belongs to cohesin protein family , playing an important role in mediating sister chromatid cohesion [54] . Solomon et al . found that the inactivation of STAG2 causes aneuploidy in human glioblastoma cell lines [55] . Lawrence et al . recently identified STAG2 as one of the 12 genes that were mutated at a substantially high frequency in at least four cancer types through examining the exome sequencing data of 4 , 742 human cancer samples across 21 cancer types [30] . Taken together , we provided statistical evidence in that hypermutation of the cancer driver genes on the inactive X chromosome may be a general feature in the female cancer genomes [52] . Further investigation on this feature is warranted .
Several previous studies showed several lines of strong biological evidences in that a single gene may shape subsequent evolution of the human cancer genome [8 , 11 , 13] . Such evidence motivated us to develop a mathematical model that can quantitatively measure a gene-gene pair to be co-mutated and highly co-expressed simultaneously in a given cancer type . Here , we proposed the gene gravity model based on Newton’s law of gravitation to study the cancer genome evolution by the systematic integration of ~3 , 000 cancer genome transcription and somatic mutation profiles from TCGA under molecular network architecture knowledge . It is worth noting that some factors , such as gene length , network topology ( e . g . connectivity ) , high mutation rate on the cancer driver genes , and high PCC value for the particular genes , may affect the performance of the gene gravity model . Longer genes would be more likely to harbor mutations , increasing the false positive rate during cancer genomic analysis [28 , 32] . We investigated the correlation of the gene aveG score with gene cDNA length collected from Tamborero et al . [56] . We removed two longest human genes ( TTN and MUC16 ) because no evidence has been found in cancer yet [28 , 32] . We observed a moderate correlation between gene aveG score and cDNA length in the 9 cancer types ( S18 Fig ) . For BRCA , the correlation is 0 . 21 between gene aveG score and gene cDNA length ( p < 2 . 2 × 10−16 ) . In addition , we recalculated the aveG score by using the average mutation density ( M/L , here M is the number of mutations for a given gene in a specific cancer type ) per base pair in each cancer type normalized by gene cDNA length ( L ) . We could reproduce the results ( S19 Fig ) , since the new results are nearly the same to those presented in S2–S10 Figs . We next examined whether the gene connectivity and gene average co-expression correlation , such as “party hub” in the network [57] , contribute to the performance of the gene gravity model . We found that gene aveG score significantly correlates with gene connectivity in all of the 9 cancer types ( S20 Fig ) . For BRCA , the correlation is 0 . 40 between the gene aveG score and gene connectivity in PIN ( p < 2 . 2 × 10−16 , F-statistics , S20 Fig ) . Thus , a gene with high connectivity may create a higher cancer genome evolution rate . Additionally , we investigated the relationship between the gene aveG and the average gene co-expression coefficient ( avePCC ) . We calculated a gene avePCC using ( ρ ) i = ∑j PCCij / n between gene i and gene j ( j belongs to the set of gene i’s interacting partners ( n ) in PIN ) based on the absolute value of PCC for each gene-gene pair . We found a moderately positive correlation between gene aveG score and its avePCC across 9 cancer types ( p < 2 . 2 × 10−16 , S21 Fig ) . Finally , we further examined whether we could reproduce the results using 4 features: high connectivity , high avePCC , gene length , and high mutation rate . For comparison , we separated genes into 3 categories based on the range of the aveG score . As shown in S22 Fig , for each of these 4 features , the distribution of aveG score cannot simply separate 3 different aveG categories: low , middle , and high groups . In a previous study , we found a positive correlation of protein connectivity with the number of nonsynonymous somatic mutations across 12 cancer types [23] . Thus , the current observation is consistent with our previous study that network-attacking perturbations due to somatic mutations occurring in the network hubs of the cancer interactome play important roles during tumor emergence and evolution [23] . There are some ultra-mutated tumor samples in various cancer types , such as UCEC or COAD . For example , a small number of tumor samples can contribute to a large proportion ( e . g . , 40% ) of total somatic mutations observed in the whole cancer cohort [29] . We removed 18 ultra-mutated tumor samples in UCEC and 31 ultra-mutated tumor samples in COAD based on a previous study [29] . We then used the remaining tumor samples to perform the same analyses . As shown in S23 and S24 Figs , we could reproduce the results , since the new results are nearly the same to those presented in Fig 2 and S2–S10 Figs . Thus , ultra-mutated tumor samples only had a minor influence on the performance of gene gravity model . There are several limitations in the current model . First , for the TCGA data , its inherent static nature gives only a single time point analysis , and we are unable to map specific genome or protein changes to the individual cells or cell populations through whole-tumor tissue analysis . Second , tumor heterogeneity and environmental factors may increase the data bias . For example , we did not find a substantial pattern indicating that the attractive gravitation for oncogenes is very stronger than that of non-oncogenes in GBM , HNSC , LUAD , or LUSC ( S12 Fig ) . One possible explanation is that environmental factors ( e . g . , smoking ) may accelerate cancer genome evolution . We separated TCGA patients into smokers and never-smokers in LUAD and LUSC , and performed the same analysis by quantifying the gene gravity model . As expected , we found that the attractive gravitation of oncogenes is significantly stronger than that of non-oncogenes for never-smokers in LUAD or LUSC ( S25 Fig ) . However , the attractive gravitation of oncogenes is marginally higher than that of non-oncogenes for smokers in LUAD or LUSC ( S25 Fig ) . Third , we used a broad context molecular network to derive the gene gravity model . However , current molecular network architectures do not completely represent the natural genetic profiles of cells . In the future , we may improve the gene gravity model in the following ways: ( i ) integrate single-cell data , including single-cell gene expression and next-generation sequencing data , to explore the dynamic features of cells and reduce the influence of tumor purity and tumor heterogeneity [58–61]; ( ii ) address cancer genetic network signatures by using large-scale genetic interaction profiles [62]; and , ( iii ) integrate panomics data resources , including the chromatin interaction network , copy number variation , proteomics , and DNA methylation profiles , to explore genomic instability more deeply and identify putative cancer driver genes [32 , 63] . Finally , we plan to use an insulated heat diffusion process implemented in a previous study [64] to consider the significance of the cancer driver genes regardless of network topology ( e . g . connectivity ) . In summary , this study reaffirms the power and value of TCGA panomic data in investigating fundamental cancer biology questions , such as somatic mutation-driven cancer genome evolution .
We downloaded the PPI data and constructed a large-context PIN from two sources: InnateDB [65] and the Protein Interaction Network Analysis ( PINA ) platform [66] . InnateDB contained more than 196 , 000 experimentally validated molecular interactions in human , mouse , and bovine models . PINA ( v2 . 0 ) is a comprehensive PPI database that integrates six high-quality public databases . We implemented three data cleaning steps . First , we defined an interaction as being high-quality if it was experimentally validated in human models through a well-defined experimental protocol . The interactions that did not satisfy this criterion were discarded . Second , we annotated all protein-coding genes using gene Entrez ID , chromosome location , and the gene official symbols from the NCBI database ( http://www . ncbi . nlm . nih . gov/ ) . Finally , duplicated or self-loop interactions were removed . In total , we obtained 113 , 473 unique interactions connecting 13 , 579 protein-coding genes ( S2 Table ) . We collected RNA-Seq data ( V2 ) from 3 , 487 tumor samples across 9 cancer types from TCGA ( http://cancergenome . nih . gov/ ) . These 9 cancer types consisted of BRCA , COAD , GBM , HNSC , KIRC , LUAD , LUSC , OV , and UCEC ( S1 Table ) . In this study , we implemented two criteria to select the genes that were expressed: ( i ) in a sample , we filtered out the genes whose mRNA expression was below the 20% of all mRNAs ordered by their expression level; and ( ii ) we further filtered out the genes that expressed in less than 20% of samples in whole expression matrix . We also extracted RNA-Seq V2 data for smokers and never-smokers in LUAD and LUSC , and for the male and female genomes in GBM from TCGA ( January 05 , 2015 ) using the R package implemented in TCGA-Assembler [67] . Finally , we calculated the Pearson Correlation Coefficient ( PCC ) for each gene-gene pair and mapped the PCC value of each gene-gene pair onto above PIN to construct 9 CePINs for the 9 cancer types ( Fig 1A ) . We collected somatic mutation profiles for 2 , 946 cancer exomes in 9 cancer types ( S1 Table ) . In total , we obtained 277 , 370 nonsynonymous somatic mutations on the protein-coding regions in ~18 , 000 genes . The details of preprocessing of mutation data are provided in Kandoth et al . [29] . We also extracted somatic missense mutations for smokers and never-smokers in LUAD and LUSC , and for the male and female genomes in GBM from TCGA ( January 05 , 2015 ) using the R package implemented in TCGA-Assembler [67] . All statistical tests were conducted using the R package ( v3 . 0 . 1 , http://www . r-project . org/ ) . The q values less than 0 . 1 were considered statistically significant . | Cancer genome instabilities , such as chromosomal instability and microsatellite instability , have been recognized as a hallmark of cancer for several decades . However , distinguishing cancer functional somatic mutations from massive passenger mutations and non-genetic events is a major challenge in cancer research . Massive genomic alterations present researchers with a dilemma: does this somatic genome evolution contribute to cancer , or is it simply a byproduct of cellular processes gone awry ? In this study , we developed a new mathematical model to incorporate the genome-wide transcription and somatic mutation profiles of ~3 , 000 tumors across 9 cancer types from The Cancer Genome Atlas into a broad gene network . We found that cancer driver genes may shape somatic genome evolution by inducing mutations in other genes in cancer . This functional consequence is often generated by the combined effect of genetic and epigenetic alterations ( e . g . chromatin regulation ) . Moreover , we provided statistical evidence that hypermutation of cancer driver genes on inactive X chromosomes is a general feature in female cancer genomes and found a putative X-inactive specific gene STAG2 in uterine cancer . In summary , this work illustrates the functional consequences and evolutionary characteristics of somatic mutations during tumorigenesis through driving adaptive cancer genome evolution . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2015 | A Gene Gravity Model for the Evolution of Cancer Genomes: A Study of 3,000 Cancer Genomes across 9 Cancer Types |
Osteoporosis is a complex disorder and commonly leads to fractures in elderly persons . Genome-wide association studies ( GWAS ) have become an unbiased approach to identify variations in the genome that potentially affect health . However , the genetic variants identified so far only explain a small proportion of the heritability for complex traits . Due to the modest genetic effect size and inadequate power , true association signals may not be revealed based on a stringent genome-wide significance threshold . Here , we take advantage of SNP and transcript arrays and integrate GWAS and expression signature profiling relevant to the skeletal system in cellular and animal models to prioritize the discovery of novel candidate genes for osteoporosis-related traits , including bone mineral density ( BMD ) at the lumbar spine ( LS ) and femoral neck ( FN ) , as well as geometric indices of the hip ( femoral neck-shaft angle , NSA; femoral neck length , NL; and narrow-neck width , NW ) . A two-stage meta-analysis of GWAS from 7 , 633 Caucasian women and 3 , 657 men , revealed three novel loci associated with osteoporosis-related traits , including chromosome 1p13 . 2 ( RAP1A , p = 3 . 6×10−8 ) , 2q11 . 2 ( TBC1D8 ) , and 18q11 . 2 ( OSBPL1A ) , and confirmed a previously reported region near TNFRSF11B/OPG gene . We also prioritized 16 suggestive genome-wide significant candidate genes based on their potential involvement in skeletal metabolism . Among them , 3 candidate genes were associated with BMD in women . Notably , 2 out of these 3 genes ( GPR177 , p = 2 . 6×10−13; SOX6 , p = 6 . 4×10−10 ) associated with BMD in women have been successfully replicated in a large-scale meta-analysis of BMD , but none of the non-prioritized candidates ( associated with BMD ) did . Our results support the concept of our prioritization strategy . In the absence of direct biological support for identified genes , we highlighted the efficiency of subsequent functional characterization using publicly available expression profiling relevant to the skeletal system in cellular or whole animal models to prioritize candidate genes for further functional validation .
The feasibility of carrying out genome-wide association studies ( GWAS ) has led to the rapid progression of the field of complex-disease genetics over the past few years . Although the GWAS approach has been successful in identifying novel candidate genes leading to new discovery of pathways that are involved in the pathophysiology of diseases , the genetic variants identified so far only explain a small proportion of the heritability for complex traits [1] . Due to the modest genetic effect size and inadequate power to overcome the heterogeneity of genetic effects in meta-analysis , true association signals may not be revealed based on a stringent genome-wide significance threshold alone [2] . In addition , the majority of the GWAS have not provided much information beyond statistical signals to understand the genetic architecture for those usually novel genes that have not been studied for a particular trait/disease before . Thus , the necessity of incorporating additional information when studying the GWAS has become apparent . Expression profiling with gene signatures of cellular models have been used to characterize gene's involvement in bone metabolism and disease processes . One such approach is parathyroid hormone ( PTH ) stimulated osteoclastogenesis and osteoblast maturation for osteoblastogenesis [3] . PTH indirectly stimulated osteoclastogenesis via its receptors on osteoblasts , which then signal to osteoclast precursors to stimulate osteoclastogenesis . Impaired osteoblastic differentiation reduces bone formation and causes severe osteoporosis in animals [4] . The TNFRSF11B/OPG gene , a well-known candidate gene for osteoporosis , is involved in osteoclastogenesis through the regulation of PTH [5] . Compared to GWAS-identified candidate genes that do not show differential expression in these cellular models , genes like TNFRSF11B/OPG with differential expression are more likely to be involved in skeletal metabolism and thus more likely to be truly associated with osteoporosis . Given that the majority of the reported genome-wide significant SNPs are in the intergenic or noncoding regions [6] , it is not clear which SNP/gene might be implicated as a causal SNP/gene . Since intergenic or noncoding SNPs do not appear to affect protein sequence , it is likely that these SNPs either are in linkage disequilibrium with the causal variants or located within the transcription regulation elements of nearby genes . The relative quantification of gene transcripts may act as intermediate phenotypes between genetic loci and the clinical phenotypes . Expression quantitative trait loci ( eQTL ) analysis in specific tissues is a valuable tool to identify potentially causal SNPs [7]–[10] . By integration of genetic variants , transcriptome , and phenotypic data , investigators have the potential to provide much-needed support to prioritize the candidate susceptibility genes identified from GWAS for further validation [11]–[13] . Previously , we conducted a pilot GWAS for osteoporosis-related phenotypes in a small subset of the Framingham study participants [14] . Osteoporosis is a skeletal disorder characterized by compromised bone strength predisposing to an increased risk of fracture . The heterogeneity of osteoporosis has both an environmental and genetic basis . Although bone mineral density ( BMD ) is frequently used in the diagnosis and prognosis of osteoporosis [15] , a growing body of evidence indicates that femoral geometry also contributes importantly to hip fracture risk [16] , [17] . Both BMD and hip geometry are strongly heritable , with heritability estimates between 50% and 85% [18] . In an attempt to identify genes that are involved in the regulation of bone health related phenotypes , genetic linkage analyses [19] , [20] , candidate gene association studies [21] and recent GWAS [22]–[27] have been used to implicate several loci and candidate genes , such as OPG/RANK/RANKL [22]–[24] , [28] , LRP5 [22] , [23] , [29] , LRP4 [23] , ESR1 [23] , [30] , VDR [31] , and SP7 [24] , [25] . However , the majority of genes that contribute to genetic susceptibility to osteoporosis remain to be elucidated . Seeking to extend these initial observations , in the current study , we first performed a large-scale GWAS analysis for BMD and hip geometry in 2 , 038 women and 1 , 531 men from the Framingham Osteoporosis Study using 550 , 000 SNPs , and then replicated the top findings in 5 , 595 women and 2 , 126 men from two independent cohorts of Caucasian individuals . We then prioritized the genome-wide association findings by utilizing publicly available experiments relevant to the skeletal system in cellular or whole animal models , and provided supportive biological information for future functional validation of their involvement in bone metabolism . The expression experiments included ( 1 ) gene signatures of a mouse embryo expression atlas and mouse cellular models of osteoblastogenesis and PTH- stimulated osteoblasts; ( 2 ) eQTL analysis in human primary osteoblasts , lymphocytes and liver tissues; and ( 3 ) likelihood-based causality model selection ( LCMS ) by integrating genetic variants , gene expression profiling , and skeletal phenotypes in inbred mice to identify candidate genes causally related to bone phenotypes . An overview of the study design is provided in Figure 1 .
Significant differences of BMD and geometry indices were found between men and women in the Framingham Study with p-values <0 . 001 ( Table S1 ) . Quantile-quantile plots of observed p-values for single SNP association tests under additive genetic effect models are shown in Figure 2 . Except for the tail ( likely comprising true associations ) , the distributions of observed p-values did not deviate from the null distribution , which rules out systematic bias due to bad genotyping or population substructure in our study samples . The estimated genome control λGC for each phenotype ranged from 0 . 99 to 1 . 02 . The regression coefficients analyzed with and without adjusting for the PCs are highly correlated ( r = 0 . 95–0 . 98 ) . Thus , we do not expect these principal components to influence our results substantially . SNPs associated with each phenotype at p-values <10−6 are listed in Table S2 . For women , the most significant association was found with neck width ( NW ) for SNP rs16965654 ( MAF = 0 . 01 ) located 13Kb away from the 5′ upstream region of the WD repeat and SOCS box-containing 1 ( WSB1 ) gene on chromosome 17q11 . 1 ( p = 4 . 15×10−8 ) . For men , the most significant association was found with neck-shaft angle ( NSA ) for SNP rs11573709 ( MAF = 0 . 23 ) located in intron 7 of the RAD23 homolog B ( RAD23B ) gene on 9q31 . 2 ( p = 2 . 37×10−7 ) . We also performed association tests by combining men and women together . The most significant association was found with NW for SNP rs16965654 ( p = 6 . 89×10−10 ) . All genotyped SNPs ( n = 431–593 for sex-specific phenotypes ) with association test p-values <10−3 in Stage I were examined for replication in the Rotterdam Study ( both men and women ) and TwinsUK Study ( women only ) . We performed meta-analyses by combining results from the Framingham Study and Rotterdam Study in men and all three cohorts in women . P-values <4 . 3×10−7 from meta-analyses are considered as genome-wide significant associations ( See statistical methods section for details ) . We listed the most significant SNP on each chromosome locus with meta-analysis p-values <10−6 in Table 1 . The most significant association for men was found with NSA for SNP rs2278729 located in the intron 4 of TBC1D8 on chromosome 2q11 . 2 ( p = 1 . 48×10−7 ) . SNP rs7227401 located in intron 4 of OSBPL1A ( 18q11 . 2 ) was found to be strongly associated with NW ( p = 4 . 22×10−7 ) in men . The most significant association for women from meta-analysis was found with LS BMD for SNP rs2062375 located in the intergenic region of TNFRSF11B and COLEC10 genes on chromosome 8q24 . 12 ( p = 2 . 68×10−11 ) . SNP rs494453 located in the intron 2 of RAP1A on chromosome 1p13 . 2 was also strongly associated with NW ( p = 2 . 80×10−7 ) . The association became more significant for SNP rs494453 when combining women and men together ( p = 3 . 6×10−8 ) . None of the above associated SNPs are exonic coding SNPs . For SNPs listed in Table 1 , no significant heterogeneity across studies was found and the p-values ( as well as regression coefficients ) were not changed with or without adjustment of body weight . The quality scores of imputed SNPs in Table 1 were >0 . 98 ( IMPUTE confidence score ) for the TwinsUK Study and >0 . 84 ( MACH variance ratio ) for the Rotterdam Study . Cis-eQTLs were analyzed for eight candidate genes located within 500 kb in four genome-wide significant loci ( Table 2 ) . All eight candidate genes were expressed in bone tissue estimated by either expressed sequence tag ( EST ) in the CGAP EST cDNA library ( Figure S1 ) or human primary osteoblast samples ( Table 2 ) . However , since transcripts were not presented on expression arrays , expression of TBC1D8 was not available in human primary osteoblast samples . P-values <0 . 005 estimated by false discovery rate ( FDR ) were considered as significant . SNP rs494453 was found to be significantly associated with transcript levels of the RAP1A gene . Allele C of rs494453 is in LD with allele A of rs3767595 ( haplotype ) . The haplotype CA was associated with lower expression of RAP1A , but higher NW ( stronger bone structure ) in women . We also performed eQTL analyses in human lymphocytes and liver tissue . Expression level of the RAP1A gene was not available for either lymphocytes or liver tissue . SNPs on chromosome 2q11 . 2 ( TBC1D8 and RLP31 ) and 8q24 . 12 ( TNFRSF11B ) loci were associated with gene expressions in lymphocytes ( Table 2 ) . The most significant eSNP was found for SNP rs2278729 ( chromosome 2q11 . 2 ) with TBC1D8 expression in lymphocytes ( p = 2 . 58×10−10 ) and liver tissue ( p<10−16 , Figure S3 ) . Allele A of rs2278729 was associated with smaller NSA in men and also with lower expression of TBC1D8 transcript . The same allele A was also associated with lower RPL31 expression in lymphocytes and was marginally significant in osteoblasts . Consistency between the direction of effect on transcript levels in lymphocytes and LS BMD was observed for TNFRSF11B at the chromosome 8q24 . 12 locus , which confirmed a previous report that increased TNFRSF11B expression levels have been shown to inhibit bone resorption [32] . A previous study also demonstrated that alleles associated with decreased BMD were associated with differential allelic expression of the TNFRSF11B in lymphocytes [22] . However , we did not observe associations of genome-wide significant SNPs in/near the TNFRSF11B gene region with TNFRSF11B expression levels in human primary osteoblasts , possible due to lack of power . We investigated the candidate genes corresponding to the genome-wide significant SNPs in 4 chromosomal regions by looking at reported gene functions ( including biological processes , canonical pathways and organism processes in human and mouse ) , microRNA targets and gene-related human diseases ( Table S3 ) . Except for the TNFRSF11B gene , there were few additional data regarding the potential biological significance of other candidate genes being involved in skeletal development and bone remodeling; therefore , we performed additional analyses on expression profiles in animal experiments ( Table 3 ) . In experiment 1 , we found that PTH negatively regulated expressions of OSBPL1A and TNFRSF11B . RPL31 , IMPACT and RAP1A genes were expressed in PTH stimulated osteoblasts , but not regulated by PTH . TBC1D8 were not expressed in PTH stimulated osteoblasts . In experiment 2 , we analyzed the differential expression of candidates during osteoblast maturation . As a quality control measure , we looked at a number of known osteoblast markers , including runt-related transcription factor 2 ( Runx2 ) , collagen type 1 , alpha 1 ( Col1a1 ) , collagen type 1 , alpha 2 ( Col1a2 ) , osteocalcin , osteopontin and osteonectin . The expected expression patterns ( differential expression during maturation ) were observed in all cases . We observed that the expression of OSBPL1A , IMPACT and COLEC10 was significantly different across a time course ( Day 4 , 5 , 6 , 8 , 16 , 25 and 30 post-differentiation ) of osteoblast development ( p<0 . 0083 ) . In the third experiment using the LCMS algorithm in the B6XC3H F2 intercross mice , we found that OSBPL1A , IMPACT , RAP1A and COLEC10 genes were predicted to be causally linked with bone phenotypes ( detailed phenotypes listed in Table S4 ) based on the evidence of significantly pleiotropic effects on trait QTL and eQTL . A total of 109 suggestive genome-wide associated regions/genes ( most significant SNP with meta-analysis 4 . 3×10−7< p-value ≤5×10−5 ) were selected based on the criteria that p-values showed nominal association in the Framingham , Rotterdam and TwinsUK studies . Among them , 16 candidate genes were prioritized with results either involving the differential expression in osteoblasts or causally linked ( LCMS algorithm ) with bone phenotypes in mice ( Table 4 ) . Among 16 prioritized candidate genes/loci , PPAP2B , GPR177 , TGFBI , DOCK1 , SOX6 and PDGFD gene expressions were regulated by PTH in osteoblasts . Significant differential expression during osteoblast development was found for GPR177 , TGFBI , SOX6 and CDH2 genes . IRX2 , TGFBI and CDH2 genes showed strong expression in the skeleton compared to 24 other subsets of organ/tissue systems of the mouse embryo . Using the LCMS algorithm in inbred mice , 12 genes were predicted to be causally linked with bone phenotypes ( detailed phenotypes listed in Table S4 ) . All of the prioritized candidate genes are expressed in bone tissues . 10 genes were found to be expressed in human bone tissue from the CGAP EST cDNA library ( Figure S1 ) and the remained genes ( HECW2 , CASR , MMRN1 , IRX2 , SOX6 and SALL1 ) were found to be expressed in human primary osteoblasts . To test the probability of our candidate genes clustering into a particular biological pathway , we performed a gene set enrichment test on 24 candidate genes ( 20 loci ) from Table 2 and Table 4 . Due to lack of biological or functional annotation , IRX2 and FBXO31 genes were excluded from analyses . We found a significant clustering ( Fisher exact test p = 1 . 65×10−4; Benjamini-Hochberg multiple testing corrected p-value = 0 . 03 ) of genes involved in adhesion of cells , including CASR , CDH2 , PPAP2B , RAP1A , TGFBI and TNFRSF11B genes . We also estimated expression abundance by number of expressed sequence tag ( EST ) sequences per 200 , 000 tags in the CGAP EST cDNA library for these 24 candidate genes . Among 48 human tissues and organs , candidate genes were expressed in bone ( 17 candidate genes ) , liver ( 22 candidate genes ) , muscle ( 18 ) and adipose tissue ( 12 ) ( Figure S1 and Figure S2 ) . Expression levels of RAP1A ( p = 2 . 51×10−4 ) , RPL31 ( p = 3 . 03×10−7 ) and TNFRSF11B ( p = 1 . 69×10−3 ) genes showed over-representation in bone ( Figure S1 ) .
In this study we performed sex-specific genome-wide association studies for BMD at the LS and FN skeletal sites as well as geometric indices of the hip in adults from the Framingham Osteoporosis Study and then replicated the top finding in two independent studies . As a result of meta-analyses on 7 , 633 women and 3 , 657 men , we discovered three novel genome-wide significant loci , including chromosome 1p13 . 2 RAP1A locus ( p = 3 . 62×10−8; NW in men and women combined ) , 2q11 . 2 TBC1D8 locus ( p = 1 . 48×10−7 , NSA in men ) and 18q11 . 2 OSBPL1A locus ( p = 4 . 22×10−7 , NW in men ) . We also confirmed TNFRSF11B gene on chromosome 8q24 . 12 to be associated with LS BMD in women only ( p = 2 . 68×10−11 ) . The RAP1A gene ( chromosome 1p13 . 2 ) was predicted to be causally linked with bone phenotypes in B6xC3H F2 intercross mice . Compared to other tissues , expression levels of RAP1A showed over-representation in human bone tissue . An eSNP ( rs494453 ) located in intron 2 of RAP1A gene was also found to be significantly associated with RAP1A gene expression in human primary osteoblasts . A marginally significant differential expression during osteoblast maturation was also found in our study . RAP1A , a GTPase that mediates calcium signal transduction , has been found to mediate activities of JnK [33] . JnK has been reported to be involved in late stage osteoblast differentiation [34] and apoptosis of osteoblasts [35] . Therefore , variants in the RAP1A gene may change the activities of JnK and then impact osteoblast maturation . Further experiments are necessary to explore the role of the RAP1A gene . Both OSBPL1A and IMPACT genes located in chromosome 18q11 . 2 region were predicted to be causally linked with bone phenotypes in mice . Expressions of both genes were found to be significantly differential during osteoblast maturation . However , only expression of the OSBPL1A gene in osteoblasts was regulated by PTH . No significant eQTL was found in this region . Given the genome-wide significant SNPs were located in the OSBPL1A gene , we still cannot rule out the involvment of the nearby IMPACT gene . In addition , an in vitro study has shown that DDIT3 over-expression enhances osteoblastic differentiation in ST-2 stromal cells , a mechanism that may involve the formation of heterodimers with C/EBP-β and the sensitization of the BMP/Smad signaling pathway [36] . IMPACT protein has found to decrease expression of mouse DDIT3 protein [37]; therefore , IMPACT may negatively regulate bone formation . We estimated the statistical power of our meta-analysis at an α-level of 10−7 . In women , the power was 62–99% and >80% for effect size ( h2 ) equal to 1% and 2% , respectively . In men , the statistical power was 35–75% and >70% for effect size equal to 1% and 2% , respectively . Inadequate statistical power seems to be one of the limitations in our study . Therefore , we prioritized 16 candidate genes/loci out of 109 suggestive genome-wide suggestive candidate genes ( 4 . 3×10−7<p≤5×10−5 ) based on the expression profiling and the LCMS modeling relevant to the skeletal system . Among 16 prioritized candidate genes/loci , PPAP2B , GPR177 , SOX6 and CDH2 genes have been reported to be involved in Wnt-signaling . CASR , TGFBI and CACNB2 genes are involved in ossification , endochondrial bone formation in cartilage and calcium ion transportation , respectively ( Table S5 ) . CASR knockout mice have demonstrated decreased bone density and abnormal bone mineralization [38] . Variants in GPR177 , SOX6 and CASR genes were associated with LSBMD in women . Variants in GPR177 and SOX6 ( 2 out of 3 above genes ) have been successfully replicated in a large-scale meta-analysis of BMD on 19 , 195 Caucasian subjects ( majority of whom were women ) with association p-values <10−9 [27] , but none of the non-prioritized candidates ( associated with BMD ) did . These results support the concept of our prioritization strategy . Candidate gene/SNP prioritization strategies by gene expression and bioinformatic databases leverage the complexity of the disease phenotypes , which offers some advantages over traditional association studies that rely on strictly p-value driven approaches . A recent study demonstrated that using functional information in published references to identify the key biological relationships between genes was able to predict the success of validation in replication genotyping [39] , which also provides additional evidence for the soundness of using biological functional relevance to prioritize candidate genes from GWAS for future validation . We exploited eSNP/eQTL in multiple human tissues . Given that ( 1 ) disease-related human tissues are often difficult to obtain for research purposes; ( 2 ) eQTL analysis requires a large sample size to reach the statistical power necessary to observe subtle changes in gene expression [40]; and ( 3 ) all of the selected candidate genes were expressed in bone tissues , we believe that performing eQTL in multiple tissues , although not replacing eQTL analysis in bone tissue , does provide complementary information . Genetic control of biological functions may be tissue-specific . Analysis of cis- eQTL in the tissue type directly relevant to the phenotype has been generally shown to be more informative than the same analysis in unrelated tissue types ( such as blood ) . However , studies have found that cis-eQTLs are conserved across tissues , when genes are actively expressed in those tissues [10] , [12] , [13] , [41]–[44] . eQTL analyses in liver , adipose , brain and muscle tissues from the same individual mice suggested that , for a gene exhibiting significant cis-eQTL associations in one tissue , 63–88% ( dependent on tissue types ) of them also exhibit cis-eQTL associations in another tissue [42] . Two recent studies , quantifying allele-specific gene expression in four human cell lines ( lymphoblastoid cell , two primary fibroblasts and primary keratinocytes ) from the same individuals , observed that only 2 . 3–10% of the mRNA-associated SNPs showed tissue-specific cis-expression across these cell lines [43] , [44] . They also found that the variation of allelic ratios in gene expression among different cell lines was primarily explained by genetic variations , much more so than by specific tissue types or growth conditions [43] . Among the highly heritable transcripts ( within the upper 25th percentile for heritability ) , 70% of expression transcripts that had a significant cis-eQTL in adipose tissue also had a significant cis-eQTL in blood cells [45] . Comparing eQTL in human primary fibroblasts , Epstein-Barr virus-immortalized B cells and T cells revealed that cell-type-shared eQTL tend to have larger effects , higher significance and to cluster tightly around the transcription start site [46] . As for bone tissue , comparing gene expression in 58 human primary osteoblast samples and 57 lymphoblastoid cell samples , despite tissues obtained from different individuals , indicated that overall , there is a large overlap in genes expressed in these two cell types , as well as the associated functional pathways [47] . 60% of the top 100 eSNP in human lymphoblastoid cells also showed associations in human primary osteoblasts , which indicated that both tissue-independent and dependent eSNP were observed in primary osteoblasts and lymphoblastoid cells [47] . Taken together this evidence suggests that if genes are expressed across tissues , their allele-specific expression can be preserved and highly correlated across tissues . Thus , the expression of a gene in liver or other non-bone tissues may not directly cause a change in bone; however , it is possible that its allele-specific expression in liver is highly correlated with allele-specific expression in bone . Because of these correlations it is possible that a gene's expression in adipose or liver can serve as surrogate markers to study the eQTL; however , the real causal relationship would be occurring in bone . It is important to note that a lack of evidence from mining publicly available gene expression experiments does not necessarily exclude a gene's involvment in skeletal metabolism , given that ( 1 ) experimental models such as osteoblastogenesis or early skeletal development , do not represent all relevant processes related to osteoporosis; ( 2 ) variation in a gene leading to disease may affect protein function but not expression; and ( 3 ) absence of association between a transcript and disease-associated SNP may be due to limited statistical power or under different environmental conditions . An inherent limitation in most of the cell line in-vitro gene expression profiling experiments , such as our PTH treated osteoblasts or Epstein-Barr virus-immortalized lymphoblastoid cell lines used to perform eQTL analysis in most of the GWAS , is that the expression profiling of the cultured cells may be varying from actual expression within in-vivo cells [10] . An additonal challenge of using available experimental data is that most of the studies performed gene profing using commercialized “genome-wide” chips , which usually have a fixed number of genes and often do not include all set of genes of given interest . Therefore , prioritization of candidate genes will be biased towards well-studied genes . Few published GWAS have addressed the potential sex-difference in genetic risks of diseases . BMD and hip geometry for men and women are known to differ , as does the prevalence of osteoporotic fractures [48] . Gender differences in the heritability of osteoporosis-related phenotypes have been reported ( reviewed in [49] ) . In the current study , few overlapping associated SNPs between men and women were found , which may be expected based on epidemiological and clinical data and may also be due to lack of power . Sex-specific associations may be due to lifestyle and environmental variation between men and women . However , it also indicates that common genetic effects for both genders may be relatively rare and therefore , larger sample sizes of men and women is needed to detect their existence . Another limitation is that we are unable to distinguish the gender-specific differential expressions , since gene expression is measured in a pooled mixture of osteoblasts from males and females , although , differentiated expression between sexes is actually less likely to occur in-vitro . In summary , our study identified three novel genome-wide significant loci and prioritized 16 genome-wide suggestive candidate genes for BMD and hip geometry traits . Beyond generating a list of top associated SNPs by statistical signals , we highlighted the efficiency of our approach to reasonably prioritize association findings by utilizing publicly available expression profiling relevant to the skeletal system in cellular or whole animal models; and to provide supportive biological information for future functional validation of their involvements in bone metabolism . Resequencing of these loci is needed to determine the causal variants and genes , along with experimental functional studies to establish their precise mechanism linked to bone health related phenotypes .
Joint analysis for results from both discovery and replication stages almost always results in greater power than analyzing discovery and replication stages separately [61] . We selected SNPs with association test p-values less than 10−3 from Stage I discovery GWAS , and replicated them using meta-analysis by combining results from the Framingham Study and two independent population-based cohorts including the Rotterdam Study and the TwinsUK Study . Since both the Rotterdam and TwinsUK studies performed whole-genome genotyping using different platforms ( Illumina platforms ) , SNP imputation was performed . Fixed effect meta-analyses were then used to estimate combined p-values . We conducted expression quantitative trait locus ( eQTL ) analysis to evaluate whether the genome-wide significant SNPs for each locus also influence transcript levels of nearby genes as a cis-effect regulator ( eSNP ) in human primary osteoblasts , lymphocytes and liver tissue . In each locus , we selected nearby genes in which the genome-wide significant SNP was located within 500 Kb in the 5′ upstream of candidate genes with the assumption that SNPs are located in ( or in LD with the variants located in ) regulation elements of candidate genes . Expression experiments in primary osteoblasts , lymphocytes and liver tissues were conducted in three different study samples . For un-genotyped SNPs , imputed SNPs ( MACH variance ratio >0 . 3 ) were used in the lymphocyte expression dataset and surrogate SNPs with LD r2≥0 . 5 were used in primary osteoblasts and liver tissue datasets . | BMD and hip geometry are two major predictors of osteoporotic fractures , the most severe consequence of osteoporosis in elderly persons . We performed sex-specific genome-wide association studies ( GWAS ) for BMD at the lumbar spine and femor neck skeletal sites as well as hip geometric indices ( NSA , NL , and NW ) in the Framingham Osteoporosis Study and then replicated the top findings in two independent studies . Three novel loci were significant: in women , including chromosome 1p13 . 2 ( RAP1A ) for NW; in men , 2q11 . 2 ( TBC1D8 ) for NSA and 18q11 . 2 ( OSBPL1A ) for NW . We confirmed a previously reported region on 8q24 . 12 ( TNFRSF11B/OPG ) for lumbar spine BMD in women . In addition , we integrated GWAS signals with eQTL in several tissues and publicly available expression signature profiling in cellular and whole-animal models , and prioritized 16 candidate genes/loci based on their potential involvement in skeletal metabolism . Among three prioritized loci ( GPR177 , SOX6 , and CASR genes ) associated with BMD in women , GPR177 and SOX6 have been successfully replicated later in a large-scale meta-analysis , but none of the non-prioritized candidates ( associated with BMD ) did . Our results support the concept of using expression profiling to support the candidacy of suggestive GWAS signals that may contain important genes of interest . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"genetics",
"and",
"genomics/gene",
"expression",
"genetics",
"and",
"genomics/complex",
"traits",
"genetics",
"and",
"genomics/genetics",
"of",
"disease",
"diabetes",
"and",
"endocrinology/bone",
"and",
"mineral",
"metabolism",
"genetics",
"and",
"genomics/bioinformatics"
] | 2010 | An Integration of Genome-Wide Association Study and Gene Expression Profiling to Prioritize the Discovery of Novel Susceptibility Loci for Osteoporosis-Related Traits |
The primary abnormality in Down syndrome ( DS ) , trisomy 21 , is well known; but how this chromosomal gain produces the complex DS phenotype , including immune system defects , is not well understood . We profiled DNA methylation in total peripheral blood leukocytes ( PBL ) and T-lymphocytes from adults with DS and normal controls and found gene-specific abnormalities of CpG methylation in DS , with many of the differentially methylated genes having known or predicted roles in lymphocyte development and function . Validation of the microarray data by bisulfite sequencing and methylation-sensitive Pyrosequencing ( MS-Pyroseq ) confirmed strong differences in methylation ( p<0 . 0001 ) for each of 8 genes tested: TMEM131 , TCF7 , CD3Z/CD247 , SH3BP2 , EIF4E , PLD6 , SUMO3 , and CPT1B , in DS versus control PBL . In addition , we validated differential methylation of NOD2/CARD15 by bisulfite sequencing in DS versus control T-cells . The differentially methylated genes were found on various autosomes , with no enrichment on chromosome 21 . Differences in methylation were generally stable in a given individual , remained significant after adjusting for age , and were not due to altered cell counts . Some but not all of the differentially methylated genes showed different mean mRNA expression in DS versus control PBL; and the altered expression of 5 of these genes , TMEM131 , TCF7 , CD3Z , NOD2 , and NPDC1 , was recapitulated by exposing normal lymphocytes to the demethylating drug 5-aza-2′deoxycytidine ( 5aza-dC ) plus mitogens . We conclude that altered gene-specific DNA methylation is a recurrent and functionally relevant downstream response to trisomy 21 in human cells .
It is now 5 decades since Down syndrome ( DS ) was first shown to result from trisomy 21 [1] , [2] , and some progress has been made toward understanding the genes that contribute to the complex array of DS phenotypes – mostly by studying the effects of the trisomy on transcriptional profiles in humans and mice and by creating transgenic and trans-chromosomal mouse models [3] , [4] . We are still far from understanding the mechanisms that underlie the complex spectrum of phenotypes in DS . Survival in DS can range from death in utero to late adulthood; cardiac defects are present in about 40% of cases , while cognitive disability is invariably present but can range from mild to severe . Additionally , there are multiple blood cell-related phenotypes including leukemoid reactions and childhood leukemias , macrocytosis with or without anemia , a markedly increased incidence of autoimmune disorders , and increased susceptibility to recurrent bacterial and viral infections [5]–[10] . All of these abnormalities must ultimately reflect the downstream responses of human cells and tissues to the chromosome 21 aneuploidy . In theory , one mechanism by which cells might respond to changes in gene dosage is altered DNA methylation . Gain of methylation at cytosines in CpG dinucleotides in promoter-associated CpG islands ( CGI's ) can enforce dosage compensation in X-inactivation , and methylation in other types of CG-rich sequences including intragenic sequences and insulator elements can affect expression and hence functional gene dosage at imprinted loci . With these simple ideas in mind we set out to ask whether gains or losses of genomic DNA methylation might occur as a downstream consequence of trisomy 21 in blood cells from adults with DS . Studies profiling mRNA expression in cells and tissues with trisomy 21 have shown that while many genes on chromosome 21 are over-expressed , subsets of genes on other chromosomes also show consistently altered expression in this background due to gene-gene interactions ( for example [11]–[15] ) . So in testing for epigenetic changes downstream of trisomy 21 it is important to examine the whole genome . Here we show that a small group of genes , distributed across various chromosomes and not over-represented on chromosome 21 , are consistently altered by recurrent gains or losses of DNA methylation in PBL of adults with DS . For a subset of these genes we find altered mRNA expression in DS versus control blood cells , and we show that this altered expression can be recapitulated by exposing normal lymphocytes to the demethylating drug 5aza-dC .
To begin to ask whether PBL from adults with DS might differ epigenetically from this same tissue in normal adults we first profiled DS and normal control samples for DNA methylation genome-wide on high density microarrays , using 2 complementary platforms: MSNP and Infinium BeadChip assays . The MSNP method adapts Affymetrix SNP arrays for methylation analysis by incorporating an initial methylation-sensitive restriction digestion [16] , and queries the methylation status of CpG dinucleotides in HpaII restriction sites roughly equally spaced along all human chromosomes in intragenic and intergenic regions ( 26 , 800 SNP-tagged loci reliably informative for CpG methylation in this experiment; see Methods ) . In contrast , the Infinium methylation assay utilizes bisulfite conversion of the genomic DNA and queries the percent methylation at each of 27 , 000 CpG dinucleotides concentrated in promoter regions of 14 , 000 human genes . We used MSNP on 250 K StyI SNP arrays to compare 5 DS PBL samples ( 4 individuals; one sampled at 2 time points 6 years apart ) to 7 normal control PBL ( 7 individuals ) . After calculating the methylation indices ( MIs; see Materials and Methods ) and carrying out non-supervised hierarchical clustering of these data we could not distinguish the DS from the control PBL samples , suggesting that trisomy 21 does not cause widespread changes in DNA methylation in PBL . However , when we analyzed the methylation values by ANOVA followed by supervised hierarchical clustering , we found small sets of candidate loci with consistent differences in methylation in DS versus normal PBL ( Figure 1 and Table S1 ) . We applied Infinium BeadChip methylation assays to a larger set of PBL DNA samples , comparing 29 individuals with DS to 20 normal controls spanning the same age range . This larger dataset included all the samples that we had run by MSNP plus additional cases and controls . Similar to the findings with MSNP , non-supervised hierarchical clustering of the methylation values ( percent methylation at each CpG queried by the array ) showed no evidence for widespread alterations in DNA methylation in DS ( not shown ) , but again ANOVA and supervised clustering produced a list of candidate differentially methylated loci ( Figure 1 and Table S2 ) . The differentially methylated loci in DS versus normal PBL from both of these microarray screens were found to be distributed on most of the human autosomes , with no specific enrichment for genes on chromosome 21 ( Tables S1 and S2 ) . The reliability of the Illumina Infinium data was shown by extremely close correlations in technical replicates ( examples of x-y plots and correlation coefficients in Figure S1 ) . To further test reliability , and to ask whether the sets of differentially methylated loci clearly distinguish between DS and normal PBL we classified samples using a logistic ridge regression and validated the robustness of the classification scheme using leave-one-out cross-validation . Ridge regression was chosen for the ability of this approach to control for the colinearity of the independent variables . Leave-one-out cross-validation demonstrated that the differentially methylated loci from the Infinium screen classified DS versus normal PBL with 100% sensitivity and specificity . Strikingly , we initially observed one apparently mis-classified sample among the 29 cases of DS ( asterisk in Figure 1 ) , but this individual proved to have mosaic trisomy 21 , with less than 50% trisomic ( +21 ) cells by karyotype . None of the other individuals analyzed in the microarray experiments showed mosaicism . This statistical approach was not suitable for the MSNP screen with the smaller set of samples , but we directly tested and validated the results for 3 differentially methylated loci from MSNP , as well as 7 loci from the Infinium screen , using bisulfite sequencing and/or MS-Pyroseq ( see below ) . As a further technical consideration , single nucleotide polymorphisms ( SNPs ) in the regions probed by the Infinium assays can in principle complicate the results for a minor subset of loci . However , the absence of common annotated SNPs detected in the candidate regions from our screen ( dbSNP; http://www . ncbi . nlm . nih . gov/ ) , the representation of several of the differentially methylated loci by multiple probes on the BeadChips and , more directly , our successful direct independent validations by bisulfite sequencing and MS-Pyroseq ( see below ) indicated that most if not all of these loci are true positives . Next we sought to validate the gene-specific differential methylation in the same DS cases and normal controls using the independent and definitive methods of combined bisulfite restriction analysis ( COBRA ) and bisulfite sequencing . These validations were successful for 10/10 loci chosen from the lists of differentially methylated genes that had passed our ANOVA and fold-change criteria in the microarray data ( 3 loci from the MSNP screen and 7 loci from the Infinium screen ) . Importantly , the bisulfite sequencing showed that for each gene the differential methylation affected not only the index CpG sites queried on the microarrays but also multiple adjacent CpG dinucleotides ( examples in Figure 2 , Figures S2 , S3 and S4 ) . We next sought to determine the frequency and specificity of differential methylation in a larger series of individuals . To this end we used MS-Pyroseq , which measures the percent methylation at multiple CpGs downstream of the sequencing primer . We applied this assay to 9 of the candidate loci . For 8 of these loci , TMEM131 , TCF7 , SH3BP2 , SUMO3 , CPT1B , CD3Z , EIF4E and PLD6 ( LOC201164 ) we found strikingly different distributions of methylation values ( percent methylation averaged over multiple contiguous CpGs ) in DS compared to normal PBL ( Table 1 and Figure 3 and Figure S2 ) . The distributions of methylation values were largely dichotomous , with only minimal overlap between DS and controls , for TMEM131 , PLD6 , EIF4E , CPT1B and CD3Z , while for TCF7 , SUMO3 , and SH3BP2 the distributions were more overlapping but nonetheless showed a clear shift in the mean values in DS versus controls . For all of these genes the inter-group differences in methylation were highly statistically significant ( p< . 0001; Table 1 ) . MS-Pyroseq for the ninth gene tested , FAM62C , revealed a wide range of methylation in both groups ( DS and controls ) but the distribution of methylation still differed significantly , albeit less strongly , between the 2 groups ( Table 1 ) . In summary MS-Pyroseq in the larger case-control series validated the differential methylation for each of these 9 candidate loci , 3 from MSNP and 6 from the Infinium assays , thus giving high confidence in the quality of the primary methylation profiling data . As methylation is known to be age-dependent for certain DNA sequences in some human tissues including the immune system [17] , for 8 of the independently validated differentially methylated genes we plotted the percent methylation in DS cases and controls as a function of age ( Figure 3 and Figure S2 ) . For TMEM131 , the average percent methylation clearly declined with age in the normal controls , but was uniformly low regardless of age in the adults with DS ( Figure 3 ) . Despite the decline with age in the controls , and our deliberate sampling of PBL from both young and elderly control individuals , the levels of methylation in this region , corresponding to a CG-rich internal promoter sequence of the TMEM131 gene , never reached the very low levels seen in DS . Similar analyses showed that for all 8 candidate genes tested the difference in methylation between DS and controls was significant both before and after adjusting for age ( Table 1 ) . As we had collected PBL samples from several of the study participants with DS at multiple time points spanning from 6 months up to 7 years , we were further able to ask whether the methylation abnormalities were stable over time in these individuals . As shown in Table S3 , the degree of methylation as determined by MS-Pyroseq was generally stable over time . Mosaicism for trisomy 21 can be found in a minor subset of individuals with DS , and this finding has been associated with less severe phenotypes . The large majority of DS cases in our series had complete trisomy 21 , but 3 cases ( one of which was run on the Infinium BeadChips and all three of which were analyzed by MS-Pyroseq ) showed moderate to high mosaicism with cells disomic for chromosome 21 constituting >15% of the leukocytes in the peripheral blood . While the rarity of these mosaic cases precluded a statistical analysis , as shown in Table S4 the cases with the greatest percentage of normal diploid cells ( high level mosaics ) showed methylation values closer to the normal range . Abnormalities of B- and T-lineage lymphocytes , either functional , numerical or both , have been reported in children and adults with DS [18]–[25] . We therefore considered whether grossly altered blood cell differential counts with normal cell type-specific variation in DNA methylation might trivially account for our findings of altered methylation in DS . We first performed automated complete blood counts for 4 of the DS blood samples , all of which showed strongly reduced methylation of the internal promoter of TMEM131 and the upstream region of TCF7 and increased methylation of the upstream portion of the SH3BP2 CGI . We found that the numbers and percentages of polymorphonuclear leukocytes ( PMNs ) and total lymphocytes were all within the normal range for our clinical laboratory ( Table S5 ) . The percentages of monocytes were slightly increased ( range in the 4 DS cases 6 . 4–11 . 9 percent; normal range 4–8 percent , Table S5 ) but as measurements of DNA methylation linearly average over all cells in a given sample , this slight increase in a minor cell population would not be sufficient to account for the altered methylation in DS PBL . We next fractionated several normal PBL samples into mononuclear cells ( PBMCs ) versus PMNs on Ficoll gradients and performed COBRA and MS-Pyroseq on the genomic DNAs . This analysis revealed cell type-specific methylation levels , but we found no evidence for differences in methylation of TMEM131 , SH3BP2 , EIF4E , or TCF7 between these 2 cell populations with a direction and magnitude that could account for the altered methylation observed in DS , even if cell numbers were altered ( Figure 2A and Figure S4 ) . Only for one of the differentially methylated genes , CD3Z , did we observe a difference in methylation in PBMC compared to PMN with a magnitude and direction that could possibly account for the observed differential methylation in DS versus normal PBL based on abnormal lymphocyte numbers . However , we were able to exclude this trivial explanation for CD3Z by showing that its CpG methylation is specifically altered in purified T-cells from DS versus controls ( below ) . Lastly , given that one of the reported findings in adults with DS is an increase in the minor sub-population of T-lineage lymphocytes with the immunophenotype of natural killer ( NK ) cells , we did a further control assessing the methylation of TMEM131 in DNA from NK cells immunopurified from normal individuals . This analysis showed that the critical region of the TMEM131 gene is in fact slightly hypermethylated in normal NK cells; a pattern opposite to the hypomethylation seen in whole blood from DS ( Figure 2A ) . Similar results excluding the possibility that the observed alterations in methylation might be trivially due to increased numbers of normal NK cells were obtained for the TCF7 and SH3BP2 genes ( Figure S4 ) . To directly assess CpG methylation within an important lymphocyte subset , we next prepared genomic DNA from T-cells immuno-purified from PBL of 12 individuals with DS and 15 control individuals , and measured SUMO3 , CD3Z and SH3BP2 promoter methylation , as well as methylation of the TMEM131 internal promoter region , by MS-Pyroseq . This procedure confirmed that all of these loci are differentially methylated between DS and normal controls , not only in whole PBL but also in the T-cell preparations , thus arguing against our findings in DS PBL being trivially due to altered numbers of ( epigenetically normal ) T-cells ( Figure S5 ) . To obtain genome wide methylation data in this cell type , we next profiled promoter methylation in T-cell DNAs from 4 normal adults and 6 adults with DS for which sufficient DNA was available , using the Infinium BeadChips . Similar to our findings using total PBL , ANOVA followed by supervised clustering of the methylation values revealed a small set of differentially methylated loci ( 140 CpGs , located in 134 different genes ) in this DS versus normal T-cell comparison ( Figure 1 and Table S6 ) . Among the 108 genes ( 118 probes ) identified as differentially methylated in our Infinium data from DS versus normal total PBL , a large subset , 17 genes ( 19 probes ) , were also found to be differentially methylated in this genome-wide analysis with the T-cell preparations . This observation of gene-specific differential methylation in purified T-cells from DS versus control individuals further supports our conclusion that the epigenetic changes reported here reflect bona fide abnormalities within specific cell types and cannot be trivially accounted for by altered percentages of the major types of leukocytes . DNA methylation in cis-acting regulatory sequences can affect gene transcription , with hypermethylation of CG-rich promoter regions causing or consolidating transcriptional repression and methylation in insulator or repressor elements sometimes causing an opposite phenomenon of increased gene expression . Total RNA was available from some , though not all , of the PBL samples in this study , and we measured mRNA transcript levels in these samples by Q-PCR . As shown in Figure 4 , isoform-specific Q-PCR revealed , as predicted , greater amounts of mRNA initiating from the internal promoter region of TMEM131 in the DS PBL samples ( in which this region is relatively hypomethylated ) compared to normal PBL ( in which this region has substantial methylation ) . In contrast , expression of the longer TMEM131 mRNA isoform initiating from the upstream CGI-associated promoter , which was uniformly unmethylated in both DS and normal PBL , did not differ between these 2 groups ( Figure S6 ) . As a second example , NPDC1 was among the genes that showed consistent promoter hypomethylation in DS compared to normal PBL by the Infinium assays ( Table S2 ) and Q-PCR for NPDC1 mRNA revealed that the DS PBL samples showed , on average , greater expression of this gene than the normal PBL samples ( Figure 4 ) . A third example was TCF7 , for which Q-PCR revealed that DS PBL samples have , on average , significantly less expression of mRNA from this gene than the normal PBL samples ( Figure 4 ) . This somewhat unexpected finding suggested that the evolutionarily conserved region 5 kb upstream of the TCF7 promoter , which is hypomethylated in many of the DS PBL samples ( Table 1 , Figure 3 and Figure S2 ) , might be acting as a negative regulatory element with greater repressive function when hypomethylated . Further insight was obtained by assessing DNA methylation directly at the upstream border of the promoter-associated CGI of this gene by MS-Pyroseq , which revealed a statistically significant tendency toward hypermethylation in DS compared to normal PBL ( Figure S7 ) . Thus individuals with DS often have substantial hypomethylation of the conserved region 5 kb upstream of the TCF7 transcriptional initiation site , and also show a significant though weaker trend toward hypermethylation of the upstream border of the TCF7 CGI , located closer to the transcription initiation site . Overall , these data are consistent with regulation of mRNA expression at these 3 loci by DNA methylation – a conclusion further supported by functional experiments using a demethylating drug ( below ) . However , in interpreting the relatively wide range of expression in the primary PBL samples it is important to take into account that the TMEM131 , TCF7 and NPDC1 genes are all known to be highly inducible in response to signaling in NK cells and probably in other lymphocyte classes ( microarray data in NCBI/GEO , accessions GDS751 and ref . [26] ) , so the net expression in a given sample is likely determined by the interaction between acute environmental signals ( cytokines and cell-cell interactions ) and the baseline methylation status of the locus . Consistent with the acute inducibility of these genes and hence the wide range of expression in primary blood samples from both DS and normal individuals , the differences in expression that we observed were highly statistically significant in the overall comparisons between DS and normal controls ( Figure 4 ) , and hence correlated strongly on average with the extent of methylation , but we could not detect strong correlations between the extent of methylation and the expression level among individuals within each group . To test more directly for a functional relationship of methylation with alterations in gene expression we examined TMEM131 , TCF7 , NOD2 , SUMO3 , CPT1B , CD3Z and NPDC1 mRNA expression in a well controlled cell culture system using the demethylating drug 5aza-dC . We exposed a proliferating T-cell line ( Jurkat ) and , more importantly , normal PBMCs isolated from fresh peripheral blood samples and expanded with a cytokine ( IL-15 ) that induces proliferation of cytotoxic T-lymphocytes and NK cells , or with a general T-cell mitogen ( phytohemaglutinin; PHA ) , to 5aza-dC for 3 days . We then prepared DNA and RNA from these cells and measured DNA methylation and mRNA expression of these 7 genes . In the absence of drug we found more robust expression of TMEM131 when the PBMCs were expanded with IL-15 , while expression of TCF7 , NOD2 and NPDC1 was higher when these cells were expanded with PHA . All 4 genes were readily detected in the proliferating Jurkat cells without cytokines . In experiments using the appropriate mitogens ( IL-15 for analyzing TMEM131 and PHA for analyzing TCF7 , NOD2 and NPDC1 ) we found that TMEM131 short isoform mRNA , NOD2 mRNA and NPDC1 mRNA levels increased , while TCF7 mRNA levels decreased , as a function of exposure to 5aza-dC , both in the Jurkat cells and in the primary PBMCs ( Figure 5 and Figure 6 ) . Also shown in Figure 6 are our independent validations of the Infinium data for NOD2 by bisulfite sequencing , which confirm the relative loss of methylation in T-cells from DS patients , compared to T-cells of normal adults in the ( non-CGI ) promoter region of this gene . These data are consistent with a functional role for DNA methylation in modulating the expression of these genes in lymphocytes ( and possibly in monocytes as well ) and for the 3 genes from the PBL screen the directions of their changes in expression upon demethylation ( increased for TMEM131 and NPDC1 and decreased for TCF7 ) match the predictions based on the differences of their average expression levels in DS versus normal PBL ( Figure 4 ) . For another two genes , NOD2 and CD3Z , our analysis of mRNA expression is still in progress but the aggregate results so far are consistent with these two genes being functionally regulated by promoter methylation . The DNA methylation data ( both genes hypomethylated in DS T-cells compared to normal T-cells ) and 5aza-dC response ( significantly increased expression of both genes on exposure of normal PHA-stimulated PBMC to the demethylating drug ) are consistent with a functional role for CpG methylation in down-modulating expression of these genes in normal T-cells ( Figure 6 and data not shown ) . Furthermore , in a small number of purified T-cell samples analyzed so far ( 7 DS T-cell and 8 normal T-cell preparations ) both of these genes are over-expressed on average in the DS T-cells compared to normal T-cells ( 2-fold for NOD2 and 2 . 5-fold for CD3Z ) . However , these Q-PCR data have shown high inter-sample variability and have not yet reached statistical significance by T-tests . In contrast to the positive data obtained for TMEM131 , TCF7 , NOD2 , CD3Z and NPDC1 , the two remaining differentially methylated genes that we tested , SUMO3 and CPT1B , have shown negative data or paradoxical correlations between CpG methylation , mRNA expression and response to 5aza-dC . For the SUMO3 gene , located on chromosome 21 , we found by Q-PCR that the mean expression level in DS PBL is about 1 . 5-fold greater than in normal PBL , thus being consistent with a simple physical gene dosage effect , with no obvious compensatory effect of the promoter hypermethylation ( data not shown ) . For the CPT1B gene Q-PCR revealed a paradoxical correlation between promoter hypermethylation and significant over-expression of CPT1B mRNA in DS PBL . This paradoxical relationship was not clarified by examining the response of CPT1B transcription to 5aza-dC , as the demethylating drug caused a slight increase , not a decrease , in its mRNA levels both in Jurkat cells and in normal PBMC ( data not shown ) .
The primary cause of DS , namely trisomy 21 , has been known since 1959 , but the pathogenesis of the diverse phenotypic features of this syndrome , not only in brain and cardiac development but also in a range of blood cell-related phenotypes including macrocytic anemia , autoimmunity , and recurrent infections , remains incompletely understood . Profiling of mRNA in cells and tissues with +21 has revealed widespread changes in gene expression , mostly small in magnitude , both for genes on chromosome 21 and for large groups of genes on other chromosomes . However sorting out the importance of any given gene has been difficult . Mice with partial trisomies , transgenic mice , and recently mice engineered to carry human chromosome 21 , are useful experimental tools for assigning or excluding roles of specific genes and regions on chromosome 21 in conferring the diverse features of DS [27] . But additional experimental approaches are needed to understand the complex genomic , cellular and tissue response to this simple chromosomal aneuploidy . Previously Chango et al . used a combination of methylation-sensitive arbitrarily primed polymerase chain reaction ( MS-AP-PCR ) and quantitation of DNA fragments to find 6 fragments that were hypermethylated in PBL from 8 individuals with DS , compared to 8 normal controls [28] . The authors suggested that the observed differences might provide a mechanism to silence constitutively over expressed genes in DS , but the methods did not allow the DNA sequence of these fragments to be determined . Here we have taken a genome-wide screening approach using 2 independent platforms that are entirely distinct in how they query methylation and are complementary and largely non-overlapping in their coverage of CpG sites . MSNP uses methylation-sensitive restriction digestion as the initial step to query the methylation status of CpG dinucleotides in HpaII restriction sites in intragenic and intergenic regions , most of which are not promoter-associated; in contrast the Infinium methylation assay utilizes bisulfite conversion of the genomic DNA as the initial step and queries the percent methylation of CpG's in promoter regions , including many CpG islands as well as a large number of non-island promoter sequences . These genome-wide microarray-based screens , with validations by independent methods in a larger series of DS cases and controls , show that there are highly recurrent gene-specific epigenetic changes in this common chromosomal disorder . Our results are from analyzing blood cells , so it is important to consider what is known about the effects of trisomy 21 in this cell lineage . There are indications from studying mosaicism over time that trisomy 21 is weakly but continually selected against in hematopoietic cells [29]–[31] . Related findings in well controlled mouse models include strongly reduced growth capacity of bone marrow stem cells in the partial trisomy Ts65Dn model of DS [32] , defects in hematopoietic progenitor cells and macrocytosis in a related partial trisomy mouse model , Ts1Cje [33] , as well as hematopoietic abnormalities in the more recently created Tc1 trans-chromosomal model [34] . Combining these observations with our current data , one possibility is that there may be biological selection over time for specific patterns of altered DNA methylation in hematopoietic stem/progenitor cells that affect net cellular proliferation in this aneuploid genetic background . Ongoing biological selection acting on stochastic variations in DNA methylation could result in altered DNA methylation , as observed here , and changes in biological properties , such as the reported functional abnormalities in NK cells [22] . In considering alternative explanations for our findings , major changes in the DNA methylation machinery are less likely , given that the microarray data show only gene-specific and not widespread alterations in DNA methylation . Immunological abnormalities are prominent in DS , and many of the differentially methylated genes in Table 1 , Tables S1 and S2 have known or predicted roles in the immune system . Among the genes that we have focused on for downstream analyses in this study several ( TCF7 , SH3BP2 , CD3Z and NOD2 ) are already known to be essential for normal lymphocyte development and function , while another group ( TMEM131 , PLD6 , NPDC1and EIF4E ) are interesting candidates for such a role . TMEM131 methylation in the internal promoter region is strikingly and consistently different in DS versus normal PBL , and this gene , encoding a transmembrane protein , has been shown to be cytokine-inducible in NK cells , together with another gene that we have studied here , NPDC1 ( microarray data in NCBI/GEO , accession GDS751; ref . [26] ) . These genes are therefore intriguing biological candidates for contributing to the pathogenesis of NK cell defects in DS . Little information is available on the function of TMEM131 but this gene was reported to be sharply up-regulated between the multipotent progenitor and pro-T cell stages of differentiation , along with only a few other genes , strikingly including another gene which we have shown here to be abnormally methylated in DS , namely the transcription factor gene TCF7 [35] , which is known to be essential for normal NK cell function [36]–[39] . While these genes are evidently co-regulated in development , they can respond oppositely to acute exposures to cytokines: expression of TMEM131 and NPDC1 in NK cells is up-regulated by IL-2/PHA while TCF7 expression is down-regulated ( microarray data in NCBI/GEO accession GDS751 [26] ) . This fact is interesting in view of the inverse abnormalities in expression of TMEM131 and NPDC1 ( increased ) compared to TCF7 ( decreased ) in DS PBL , which is paralleled by the inverse responses of these genes to DNA demethylation in our experiments using 5aza-dC . Considering the possible functions of some of the other differentially methylated genes , NOD2 encodes a pathogen recognition receptor that is often studied in monocytes and macrophages but it is also expressed by CD34+ hematopoietic stem/progenitor cells and FOXP3-positive T regulatory cells [40] . SH3BP2 codes for a pleckstrin homology domain- and Src homology 2 ( SH2 ) domain-containing adaptor protein that is preferentially expressed in hematopoietic tissues including macrophages , NK , T- , and B-cells . It is involved in leukocyte signaling downstream Src/Syk-kinases and plays a crucial role in signaling during cell differentiation [41] . PLD6 , encoding a member of the phospholipase-D family , has not yet been studied for its role in lymphocyte physiology but other phospholipase-D family members are known to be crucial for signaling downstream of the T-cell receptor [42] . EIF4E , encoding a translational initiation factor that is rate-limiting for expression of numerous proliferation-related cellular proteins , is up-regulated during T-cell activation and promotes lymphocyte chemotaxis [43] , [44] . SUMO3 is one of a family of small sumo proteins that modify the activities of other cellular proteins by post-translational sumoylation – a process that is known to affect lymphocyte physiology including regulation of immunoglobulin production by B-cells [45] and mitogenesis and cytokine production in T-cells [46] , [47] . The biological role of the protein encoded by the NPDC1 gene is not yet known , and knockout mice lacking this gene did not show an obvious developmental phenotype [48] . Based on our findings of altered methylation and expression of this gene in DS PBL , and the observation of its induction during NK cell activation noted above , it would be interesting to study immune system function in the Npdc1-deficient mice . Tables S1 and S2 contain additional examples of differentially methylated genes encoding cytokines , receptors and transcription factors that also warrant examination for roles in the normal immune system and in the immunological abnormalities associated with DS . In summary , our findings show that recurrent gene-specific alterations in CpG methylation are a stereotypical cellular response to trisomy 21 , with functional consequences in gene regulation . Interesting gene candidates for the immune dysfunction in DS are already emerging from these data , and as insights from studying DS as a model system have often shed light on physiological mechanisms in the general population it will also be important to dissect the roles in the normal immune system of the genes from our screen . Additional screens , including analysis of other tissues such as heart and brain , will be useful for pinpointing loci that are recurrently altered by gains or losses of DNA methylation in other cell types that contribute to key aspects of DS such as cardiac defects and cognitive disability . A more general corollary of our findings , beyond trisomy 21 , is that there may be recurrent and predictable epigenetic consequences of other chromosomal copy number aberrations - for example in several types of human cancers , such as leukemias , Wilms tumors , and sarcomas , that frequently have simple aneuploid karyotypes .
This study was approved by Institutional Review Boards of the New York State Institute for Basic Research and Columbia University Medical Center . Participants with DS were ascertained through the New York State developmental disability service system as well as agencies in New Jersey , Connecticut and Northern Pennsylvania and have been assessed comprehensively including full medical chart reviews . The participants were recruited through responsible state and private service agencies , who contacted the participant's families or correspondents for permission for us to recruit . Informed consent was provided by either a parent or correspondent , and assent was obtained from the participant . The distribution of age , level of intellectual disability and residential placement did not differ between those participating and those who refused . Age-matched control participants were laboratory volunteers and participants in the Washington Heights-Inwood Community Aging Project who gave informed consent for genetic studies . Confirmation of trisomy 21 by G-banded karyotypes was available for 98% of the study participants with DS , with 100% concordance between cytogenetics and the clinical diagnosis of DS . Of those karyotyped , the large majority had complete trisomy 21 . However , 7 cases exhibited low level mosaicism with most of the cells having trisomy 21 and less than 15% of the cells showing a diploid chromosome complement , 3 cases showed higher level mosaicism with greater than 15 percent of cells having 46 chromosomes ( disomic for chromosome 21 ) and six cases presented with Robertsonian translocations , which in each case produced complete trisomy for the euchromatic region of chromosome 21 in all cells . MSNP on Affymetrix 250 K StyI arrays was carried out essentially as previously described [16] , [49] , [50] . Each biological sample ( total peripheral blood leukocyte DNA ) was analyzed by hybridizing the arrays with genomic representations ( probes ) made according to the Affymetrix protocol , with the following pre-digestions of the genomic DNA as the first step in the procedure: StyI ( S ) , StyI+HpaII ( SH ) , StyI+MspI ( SM ) . All other steps subsequent to the genomic pre-digestions were according to the Affymetrix protocol . The S , and SH representations were prepared and hybridized in duplicate for each biological sample; the SM representations were single for each sample . Infinium Human Methylation27 BeadChip ( Illumina ) assays , based on bisulfite conversion of genomic DNA followed by primer extension on the BeadChips to query the methylation status of defined CpG dinucleotides , were performed according to the protocol from the manufacturer . The MSNP data ( . cel files ) were processed in dChip ( [51]; http://biosun1 . harvard . edu/complab/dchip/ ) by normalization , model-based expression , and chromosome analysis . We assigned a numerical ploidy of 2 to the S arrays from the normal PBL samples , leaving the ploidy field blank for all other arrays . This strategy allowed us to visualize , using the chromosome view in dChip , the methylation status of HpaII sites flanking a given SNP-tagged locus as the extent of reduction in signal intensity in the SH representations , compared to the S representations . As MspI is the methylation-insensitive isoschizomer of HpaII , the signal intensities observed in the SM representations allowed us to determine the reliability of the Class 2 SNPs ( those with adjacent HpaII sites thus informative for methylation status [16] , [49] , [50] ) , with reliable Affymetrix probe sets indicated by strong reduction in signal in SM compared to S . For the 26 , 800 Class 2 loci with SMav/Sav<0 . 5 we calculated the methylation index ( MI ) as the fractional preservation of intensity in SH compared to S . Similar lists of candidate differentially methylated genes were obtained when we first subtracted SMav as background and then calculated the methylation index . The Infinium BeadChip data were processed using Genome Studio software , which calculates the percent methylation at each CpG queried by the arrays . The numerical values for methylation index ( MSNP ) and percent methylation ( Infinium ) were imported to dChip as external data and analyzed by ANOVA and supervised hierarchical clustering after removing all probes for genes on the X or Y chromosome , and applying fold-change and absolute difference criteria ( Results and Figure 1 legend ) . To statistically validate the Infinium data we classified DS versus normal PBL using the % methylation of the differentially methylated loci in Table S2 using a logistic ridge regression . The ridge parameter was set to 10−8 . Leave-one-out cross-validation was used to demonstrate that the classifier was not over trained to our particular test samples . The ridge regression , sensitivity , and specificity calculations were performed using Weka 3 . 4 . Genomic DNA , 0 . 6 to 1 microgram , was bisulfite-converted using the EpiTect Bisulfite Kit ( QIAGEN ) according to the instructions of the manufacturer . Sequences including or adjacent to the index SNPs or Infinium CpG dinucleotides were amplified by PCR , using primers designed in MethPrimer [52] . PCR conditions , primer sequences , and corresponding unconverted genomic sequences are in Table S7 . For COBRA we identified restriction sites in the converted sequences that differed according to methylation status of specific CpG dinucleotides and we digested the bisulfite PCR products with these enzymes followed by electrophoresis on 1 . 5% agarose gels . For bisulfite sequencing the PCR products were cloned using the TopoTA Cloning System ( Invitrogen ) and >12 plasmids sequenced for each gene in a given individual . MS-Pyroseq was performed by bisulfite converting genomic DNA samples , followed by PCR with gene-specific primers ( designed in MethPrimer ) and Pyrosequencing of the resulting PCR products at EpigenDx ( Worcester , MA ) using a Qiagen PSQ instrument . The methylation indices from MS-Pyroseq were calculated as the average percent methylation of ≥8 successive CpG dinucleotides between the primers ( Table S7 ) . Q-PCR was performed using a 7300 Fast Real-Time PCR System ( Applied Biosystems ) . Reactions were performed in triplicate in 96-well optical reaction plates . Each reaction contained cDNA reverse transcribed from 5 ng total RNA , 1X Power SYBR Green PCR master mix ( Applied Biosystems ) and 0 . 2 µM of each specific primer pair , which were designed using online D-Lux ( Invitrogen ) or Primer Express 3 . 0 software ( Applied Biosystems ) . The thermal cycling conditions were primer annealing at 50°C for 2 min and an initial denaturation for 10 min at 95°C , followed by 40 cycles of 15 s at 95°C for denaturation and 1 min at 60°C for annealing and extension . The relative expression level of a target gene in a particular sample was calculated by the delta-CT method as described [53] . NK cells were purified from human blood to >90% purity using immunomagnetic beads as previously described [54] . T-cells were isolated from blood of adults with DS and normal adult controls to >80% purity using a RosetteSep Kit ( Sigma ) according to the manufacturer's instructions . | Down syndrome ( DS; trisomy 21 ) is caused by the gain of a single extra chromosome 21 . However , the mechanisms by which this extra chromosome produces the medical abnormalities seen in DS , including not only mental retardation but also susceptibility to autoimmune diseases and recurrent infections , are still not understood . DNA methylation is a mechanism that might contribute to these abnormalities . To test this possibility , we profiled DNA methylation in white blood cells from adults with DS and normal controls and found recurrent abnormalities of gene methylation in DS , with several of the differentially methylated genes having roles in blood cells . Among the genes with hypo- or hyper-methylation in white blood cells or purified T-lymphocytes from adults with DS , compared to these same types of cells from normal adults , were TMEM131 , TCF7 , CD3Z , SH3BP2 , EIF4E , SUMO3 , CPT1B , NOD2/CARD15 , NPDC1 , and PLD6 . Several of these genes showed not only different methylation but also different expression in DS versus control blood cells , which was recapitulated by exposing normal white blood cells to a demethylating drug . These findings show that altered DNA methylation of a specific group of genes is a fundamental cellular response to the gain of an extra chromosome 21 in humans . The abnormally methylated genes identified here may contribute to immune system abnormalities in people with DS . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"genetics",
"and",
"genomics/chromosome",
"biology",
"genetics",
"and",
"genomics/genetics",
"of",
"disease",
"genetics",
"and",
"genomics/epigenetics",
"genetics",
"and",
"genomics",
"genetics",
"and",
"genomics/medical",
"genetics"
] | 2010 | Altered DNA Methylation in Leukocytes with Trisomy 21 |
The HIV-1 gp120-gp41 complex , which mediates viral fusion and cellular entry , undergoes rapid evolution within its external glycan shield to enable escape from neutralizing antibody ( NAb ) . Understanding how conserved protein determinants retain functionality in the context of such evolution is important for their evaluation and exploitation as potential drug and/or vaccine targets . In this study , we examined how the conserved gp120-gp41 association site , formed by the N- and C-terminal segments of gp120 and the disulfide-bonded region ( DSR ) of gp41 , adapts to glycan changes that are linked to neutralization sensitivity . To this end , a DSR mutant virus ( K601D ) with defective gp120-association was sequentially passaged in peripheral blood mononuclear cells to select suppressor mutations . We reasoned that the locations of suppressors point to structural elements that are functionally linked to the gp120-gp41 association site . In culture 1 , gp120 association and viral replication was restored by loss of the conserved glycan at Asn136 in V1 ( T138N mutation ) in conjunction with the L494I substitution in C5 within the association site . In culture 2 , replication was restored with deletion of the N139INN sequence , which ablates the overlapping Asn141-Asn142-Ser-Ser potential N-linked glycosylation sequons in V1 , in conjunction with D601N in the DSR . The 136 and 142 glycan mutations appeared to exert their suppressive effects by altering the dependence of gp120-gp41 interactions on the DSR residues , Leu593 , Trp596 and Lys601 . The 136 and/or 142 glycan mutations increased the sensitivity of HIV-1 pseudovirions to the glycan-dependent NAbs 2G12 and PG16 , and also pooled IgG obtained from HIV-1-infected individuals . Thus adjacent V1 glycans allosterically modulate the distal gp120-gp41 association site . We propose that this represents a mechanism for functional adaptation of the gp120-gp41 association site to an evolving glycan shield in a setting of NAb selection .
The HIV-1 envelope glycoprotein ( Env ) complex comprises a trimer of gp120 subunits in non-covalent association with a trimer of transmembrane gp41 subunits and mediates viral attachment , membrane fusion and viral entry ( for review see [1] , [2] ) . Within gp120 , 5 conserved regions ( C1–C5 ) alternate with 5 variable regions ( V1–V5 ) . The conserved regions largely form the gp120 core comprised of inner and outer subdomains that are bridged by 4 antiparallel β-strands ( the bridging sheet ) , whereas the variable regions form external solvent-exposed loops [3] , [4] , [5] , [6] , [7] , [8] . gp120 is anchored to the viral envelope by the trimeric transmembrane/fusion glycoprotein , gp41 . The ectodomain of gp41 comprises an N-terminal fusion peptide linked through N- and C-terminal α-helical heptad repeat sequences ( HR1 and HR2 , respectively ) to a C-terminal membrane anchor and cytoplasmic tail . A central disulfide-bonded loop region or DSR joins HR1 to HR2 ( Fig . 1A , B ) . The membrane fusion and viral entry function of gp120-gp41 involves conformational changes that are triggered by receptors . CD4 ligation is believed to reorganize V1V2 and V3 to expose a binding site for the chemokine receptors CCR5 and CXCR4 , which function as fusion cofactors [3] , [4] , [5] , [6] , [9] , [10] , [11] , [12] . The V3 loop mediates important contacts with the negatively charged N-terminal domain and extracellular loop 2 of CCR5 and CXCR4 and determines the chemokine receptor preference of HIV-1 isolates . In a virion context , CD4 binding causes an “opening up” of the gp120 trimer due to outward rotation and displacement of gp120 monomers [10] , [12] . gp120-receptor interactions cause gp41 to transition from a dormant metastable structure into a fusion active state [1] , [2] , [13] , [14] . Structural transitions in gp41 that are associated with fusion function include the formation of a “prehairpin intermediate” structure wherein a triple-stranded coiled coil of HR1 segments provides a binding surface for the HR2 , while studies with synthetic peptides and glycoprotein mutants indicate that the fusion peptide inserts into the target membrane [15] , [16] , [17] , [18] , [19] . Antiparallel HR1-HR2 interactions form a 6-helix bundle which apposes the N- and C-terminal membrane inserted ends of gp41 , and the associated viral and cellular membranes , leading to merger and pore formation [20] , [21] , [22] , [23] , [24] . How conformational signals are transmitted between receptor-bound gp120 and gp41 to trigger the refolding of gp41 into the fusion-active state is being elucidated . A gp120-gp41 association site formed by the terminal segments of C1 and C5 of gp120 and the central DSR of gp41 [25] , [26] , [27] , [28] may play an important role in this process as mutations in the DSR can inhibit CD4-triggered gp41 prehairpin formation and the initial hemifusion event [29] . Furthermore , the introduction of Cys residues to C5 and to the DSR generates an inactive disulfide-linked gp120-gp41 complex that is converted to a fusion-competent form by reduction [30] , [31] . These findings implicate the C1–C5-DSR synapse in maintaining gp120-gp41 in the prefusion state and in subsequent transmission of fusion activation signals emanating from receptor-bound gp120 . The terminal C1 and C5 gp41-contact regions project ∼35-Å from a 7-stranded β-sandwich at the base of the gp120 inner domain [8] . This β-sandwich appears to also play an important role in conformational signalling between gp120 and gp41 by linking CD4-induced structural changes in 3 structural layers of gp120 that emanate from the β-sandwich to the activation of gp41 [32] . Understanding how conserved functional determinants of the HIV-1 glycoproteins tolerate or adapt to the rapid evolution of other Env regions is important for their evaluation and exploitation as potential drug and/or vaccine targets . For example , resistance to a novel low molecular weight fusion inhibitor , PF-68742 , is conferred by mutations in the DSR implicating this gp41 region as an inhibitor target [33] . Neutralizing antibodies exert strong evolutionary pressures on Env that can result in an increase in the number and/or a change in the position of potential N-linked glycosylation sites ( PNGSs ) that modify NAb-Env interactions [34] , [35] . V1V2 is a key regulator of neutralization resistance , which generally correlates with its elongation and acquisition of PNGSs [34] , [36] , [37] , [38] , [39] , [40] , [41] , [42] , [43] , [44] . Previously , we identified the C1–C5-DSR association site as a conserved determinant that exhibits structural and functional plasticity . This idea is based on the finding that whereas the gp120-gp41 association function and sequence of the DSR is conserved , the contribution of individual DSR residues to gp120 anchoring and membrane fusion function varies among HIV-1 strains and is controlled by V1V2 and V3 [28] . We proposed that this plasticity enables the maintenance of a functional glycoprotein complex in a setting of host selective pressures such as NAbs that drive the rapid coevolution of gp120 and gp41 . In the present study , we reveal a structural mechanism whereby the conserved gp120-gp41 association site adapts to changes in the glycan shield . Forced viral evolution was used to examine how the gp120-gp41 complex adapts to overcome a mutation ( K601D in the DSR ) that causes defective gp120-gp41 association . The defective phenotype was restored by second site mutations leading to the loss of PNGSs at V1 positions 136 and 142 , the former operating in conjunction with L494I in C5 of gp120 and the latter requiring D601N in the DSR . The PNGS mutations appeared to exert their suppressive effects by altering the dependence of gp120-gp41 interactions on Leu593 , Trp596 and Lys601 within the DSR . The PNGS mutations were also linked to an increase in the sensitivity of HIV-1 pseudovirions to neutralization by the glycan-directed NAbs 2G12 and PG16 . Our data implicate V1 glycans linked to neutralization susceptibility in the allosteric modulation of the gp120-gp41 association site . We propose that this represents a mechanism for maintaining gp120-gp41 function in a setting of NAb-driven Env evolution .
In this study , we subjected the gp120-gp41 association defective DSR mutant HIV-1AD8-K601D to serial PBMC passage in order to select suppressor mutations . We reasoned that the locations of suppressors would point to structural elements that are functionally linked to the DSR , thereby shedding light on its role in Env function . We first confirmed that the K601D mutation promoted the shedding of gp120 into the culture supernatant of 293T cells transfected with pcDNA3 . 1-AD8env , as determined by radioimmunoprecipitation ( Fig . 1C ) . Consistent with this finding , K601D-mutated HIV-1AD8 virus particles were largely devoid of gp120 ( Fig . 1D ) . We note the presence of gp160 in the pelleted virion preparations . It likely represents one of a number of non-native Env forms that have been observed in virions and have been implicated as immune decoys [45] , [46] , [47] , [48] , [49] , [50] . The gp120 shedding defect did not appear to be associated with a precursor processing defect as Western blotting with the gp41-directed C8 monoclonal antibody ( mAb ) , indicated that similar amounts of gp41 were derived from gp160 for both WT and K601D ( Fig . 1E ) . Cell-to-cell fusion function was blocked by K601D at sub-saturating Env conditions ( Fig . 1F ) , while HIV-1AD8 viral replication over 10 days in phytohemagglutinin-stimulated peripheral blood mononuclear cells ( PBMCs ) from independent donors was severely attenuated by the mutation ( Fig . 1G ) . The restoration of HIV-1AD8-K601D replication competence was observed with sequential passage of cell-free virus in independent PHA-stimulated PBMC cultures ( Fig . 1G , cultures P0 and P2 ) . In order to identify suppressors of K601D , the env region was PCR-amplified from genomic DNA isolated from infected cells at days 20 , 30 , 40 and 50 for culture P0 , and days 30 and 50 for culture P2 . The PCR products were cloned into the pΔKAD8env expression vector for DNA sequencing . The K601D mutation was retained in P0 clones isolated at days 20 and 30 ( FIG . 2A ) , when replication was not observed in the mutant virus culture ( FIG . 1G , P0 ) . The emergence of replication-competent virus at day 40 in the P0 culture correlated with the appearance of L494I in C5 of gp120 ( 6/6 clones ) , together with either a D601N pseudoreversion ( 3/6 clones ) , and/or T138N , which abolishes a conserved PNGS at position 136 ( N136VT ) within V1 of gp120 ( 3/6 clones ) . By day 50 , when mutant viral replication approached that of WT , the T138N/L494I/K601N triple mutation was present in 42% of clones . A single clone ( P0 . D50 . 10 ) contained S144N , which ablates the nearby PNGS at N142SS . In an independent PBMC long-term culture ( P2 ) , the replication competence of the mutant virus approached that of WT by the 3rd passage ( Fig . 1G , P2 ) . Deletion of residues N139INN ( ΔN139INN ) from V1 , that results in the loss of overlapping PNGSs at positions 141 and 142 ( N141N142SS ) , was observed in 3/14 P2 day-30 clones , while 2/14 possessed mutations at Thr138; K601N was also observed in 50% of P2-day 30 clones ( Fig . 2B ) . The dominant genotypes in P2 day-50 clones were ΔN139INN together with K601N , and K601N alone . A 2-residue N142S143 deletion was observed in 1 day-50 clone ( P2 . D50 . 141 ) , which also ablates the overlapping PNGSs at 141 and 142 . Interestingly , the L494I mutation was not observed in P2 clones . These data indicate that PNGSs in V1 are functionally linked to the gp120-gp41 association synapse . To determine how replication was restored to the K601D virus in the long-term PBMC cultures , we reconstructed the dominant P0 and P2 genotypes in the context of the pAD8 proviral clone and examined viral replication in two independent PBMC donors . The P0 mutations , L494I within C5 and K601N within the DSR , partially restored replication in the PBMCs of one donor but not in the second , apparently less permissive donor ( Fig . 3A and B; P0 panels; Table S1 ) . The T138N/L494I/K601D and T138N/L494I/K601N genotypes exhibited near-WT replication kinetics in PBMCs from both donors . These data suggest that T138N , which ablates the PNGS at Asn136 within V1 , and L494I act synergistically to suppress K601D ( and K601N ) . The analysis of P2 genotypes in PBMC cultures indicated that the N139INN deletion in V1 was not sufficient to suppress K601D , but its combination with K601N led to near-WT levels of viral replication ( Fig . 3A , B; P2 panels ) . The pΔKAD8env expression vector was next employed to further dissect the functional linkages between position 601 within the DSR , the Asn136 and Asn141/142 PNGS mutations in V1 and the L494I mutation in C5 in gp120-gp41 association and cell-cell fusion assays . The results presented in Fig . 3C indicate that K601D and K601N DSR mutations promote shedding of gp120 into culture supernatants of transfected cells by ∼6 . 8- and 4 . 4-fold with respect to WT , respectively , while cell-cell fusion activity was reduced by ∼85–90% ( Fig . 3D; Table S1 ) . These data indicate that the K601D and K601N DSR mutations inhibit cell-cell fusion function by destabilizing the gp120-gp41 complex . However , the effect of K601N on Env function in the context of virus replication appears to be donor dependent ( see Fig . 3A and B ) . It may be that lower receptor numbers and/or alternate coreceptor post-translational modifications on target cells , or differences in target cell population numbers [51] , [52] , [53] , [54] in the case of the latter donor's PBMCs do not enable the fusion-activation threshold to be reached for K601N Env . A general correlation between the restoration of cell-cell fusion activity and gp120-gp41 association was observed for the P0 combination mutants . Thus L494I/K601D exhibited ∼50% of WT cell-cell fusion activity with improved gp120-gp41 association , whereas T138N/L494I/K601D and T138N/L494I/K601N were functionally similar to the WT . The phenotype of T138N/K601D was an outlier as gp120-gp41 association was improved by ∼2-fold with respect to K601D without a restorative effect on Env fusion function . These data indicate that T138N and L494I act cooperatively to restore gp120-gp41 association to K601D with concomitant restoration of membrane fusion function and viral replication competence . T138N-containing gp120 molecules migrated to lower molecular weight positions with respect to WT in reducing SDS-PAGE , consistent with loss of the glycan at Asn136 ( Fig . 3C ) . The ΔN139INN/K601D and ΔN139INN/K601N P2 clones exhibited 74–77% of WT fusion activity ( P<0 . 05 ) with only partial restoration of gp120-gp41 association , even though only the latter clone was competent for replication in PBMCs ( Figs . 3A–D , Table S1 ) . The suppression of the K601N fusion and replication phenotype by ΔN139INN is therefore not dependent on the full restoration of gp120-gp41 association or on the L494I C5 mutation . gp120 molecules with the ΔN139INN mutation migrated faster in SDS-PAGE than non-ΔN139INN-containing Envs , again consistent with the loss of glycan ( Fig . 3C ) . The ΔN139INN mutation disrupts the overlapping N-linked glycosylation sequons: Asn141-Asn142-Ser-Ser . Such overlapping N-linked glycosylation sequons are observed in V1 of a subset of HIV-1 strains ( http://www . hiv . lanl . gov/content/sequence/NEWALIGN/align . html#comp ) , although the position varies . Combining S144N ( a PNGS mutation observed in clone P0 . D50 . 10 ) with K601N resulted in almost identical fusion activity to ΔN139INN/K601N ( Fig . 3D ) , suggesting that loss of the Asn142 glycan can substitute for the N139INN deletion . Scanlan and coworkers [55] , [56] reported that gp120 derived from virions produced in 293T cells and PBMCs is composed of oligomannose type glycans , whereas monomeric gp120 produced in 293T cells contains predominantly complex type glycans in addition to the oligomannose type . It was therefore important to determine whether the presence of complex type glycans on cell-surface expressed Env was influencing the results of the cell-cell fusion and gp120-gp41 association assays . Kifunensine , an α-mannosidase I inhibitor that blocks the synthesis of Man5GlcNac2 , the precursor of hybrid and complex-type glycans [57] , was used to restrict Env glycosylation to the oligomannose type . The culturing of Env-expressing 293T cells in increasing concentrations of kifunensine did not appear to affect the efficiency of gp160 processing to gp120 and gp41 , however , the resultant gp120 migrated faster than its counterpart expressed in the absence of kifunensine . Furthermore , the gp120 and gp41 bands had a more focused appearance relative to the corresponding glycoproteins expressed under control conditions ( Fig . S1A ) . These data are consistent with homogeneous oligomannose type Env glycosylation due to the action of kifunensine [58] . We next examined the effects of 10 µM kifunensine on the cell-cell fusion activities of the T138N/L494I/K601N and ΔN139INN/K601N revertant Envs . Figure S1B shows that the presence of kifunensine leads to a general elevation in the fusion activities of the 4 constructs tested . Importantly , the presence of kifunensine did not affect the relative fusion activities of the T138N/L494I/K601N and ΔN139INN/K601N revertant Envs with respect to wild type and K601D . These data were reflected in the gp120-gp41 association assay where the shedding phenotype of K601D was restored to wild type and near-wild type levels , respectively , in T138N/L494I/K601N and ΔN139INN/K601N , in both the presence and absence of kifunensine ( Fig . S1C ) . Overall , these data indicate that the fusion and gp120-gp41 association phenotypes of the cell-expressed revertant Envs are not influenced by the presence of complex or hybrid-type glycans . We next asked whether the V1 PNGS mutations restored function via a functional link to Asp or Asn at position 601 in the DSR or whether a generalized enhancement in Env function could explain the restored phenotypes . Figures 3E and F indicate that T138N and ΔN139INN do not increase the cell-cell fusion activity of surface expressed Env or the infectivity of Env-pseudotyped luciferase reporter virus when introduced to the WT background . We also noted that the L494I mutation in a WT background did not substantially increase cell-cell fusion or gp120-gp41 association ( data not shown ) . These data are consistent with specific functional crosstalk between the V1 glycans , the Asp601 or Asn601 in the DSR and position 494 in C5 . Finally , we asked if the V1 PNGS mutations at 136 and 142 restored function via a specific link to position 601 in the DSR , or whether deletion of any of the 6 PNGSs in V1V2 of the AD8 strain ( Fig . 4A ) could restore functional defects related to a DSR mutation . To this end , we introduced T138N , S143N , S144N , S158N , N160Q and S188N mutations to the WT and K601N pΔKADenv vectors and determined their effects on cell-cell fusion , glycoprotein processing and gp120-gp41 association . Figure 4B indicates that in a WT Env context , T138N and S188N did not affect cell-cell fusion , S143N , S144N and N160Q led to small but significant functional enhancements , whereas S158N blocked fusion completely . Even though , most of the V1V2 glycan mutants were fusogenic on a WT background , only T138N and S144N restored function to K601N . A Western blot of ΔKADenv-transfected 293T cell lysates with the gp41-specific mAb , C8 , confirmed that the mutant glycoproteins were expressed and processed to gp41 at levels that were close to the WT ( Fig . 4C ) . Pulse-chase biosynthetic labeling followed by radioimmunoprecipitation with pooled IgG derived from HIV-1-infected individuals ( HIVIG ) indicated WT-like gp120-gp41 association for T138N , S143N , S144N , N160Q and S188N and a shedding phenotype for the fusion-defective S158N ( Fig . 4D ) . Near-WT levels of gp120-gp41 association were observed for the fusion-competent T138N/K601N mutant , whereas S144N/K601N , which is also fusion competent , exhibited a mild gp120 shedding phenotype . By contrast , improvements in gp120-gp41 association did not follow the combination of K601N with S143N , S158N , N160Q or S188N , consistent with their lack of fusion activity . Overall , these data are consistent with a specific functional linkage between the DSR and the Asn136 and Asn142 glycosylation sites in V1 . The effects of reversion-associated mutations on gp120-CD4 binding , sensitivity of cell-cell fusion and infectivity to sCD4 inhibition , CCR5 utilization and sensitivity to the HR2-based fusion inhibitor peptide , C34 , were next examined in order to further elucidate the mechanism whereby function was restored in the context of a mutated DSR . CD4 binding curves were generated by incubating a constant amount of biosynthetically labelled WT and mutated monomeric gp120 with serial dilutions of sCD4 , coimmunoprecipitation of gp120-sCD4 complexes with mAb OKT4 , followed by SDS-PAGE and densitometry of gp120 bands . Similar CD4 binding curves were observed for WT , T138N/L494I and ΔN139INN gp120 mutants ( Fig . 5A ) . The sCD4 EC50 for WT , T138N/L494I and ΔN139INN gp120 was ∼0 . 5 nM , which approximates the published affinity range of sCD4 for monomeric gp120 [59] . These data indicate that the glycosylation site mutations in V1 and L494I in C5 did not alter the CD4-binding ability of monomeric gp120 . We next compared the sensitivity of T138N/L494I/K601N- , ΔN139INN/K601N- and WT-Env-mediated fusion to inhibition with sCD4 . The Env-expressing 293T cells were incubated with a sCD4 dilution series for 3 . 5 h prior to an 8-h coculture with CD4-plus-CCR5-expressing BHK21 targets harboring a luciferase reporter . Figure 5B indicates that WT and T138N/L494I/K601N have almost identical sensitivities to sCD4 with IC50s of ∼60 µg/ml , which are comparable to previously published values for both T cell-line adapted and primary HIV-1 Envs [60] . The ΔN139INN/K601N inhibition curve , however , was shifted by ∼1log2 to the left , indicating that this Env is slightly more resistant to sCD4 , even though monomeric gp120 containing the ΔN139INN mutation had similar sCD4 binding characteristics to WT and T138N/L494I . We sought to recapitulate the cell-cell fusion data in a virus infection system . For this , we used TZM-bl luciferase reporter cells and the CD4-IgG2 fusion protein wherein the Fv regions of IgG have been replaced by domains 1 and 2 of CD4 . The tetrameric nature of CD4-IgG2 leads to a reduction in IC50 relative to monomeric sCD4 [60] . Complete inhibition of virus infectivity was achieved with 15 µg/ml of CD4-IgG2 with WT and T138N/L494I/K601N exhibiting similar sensitivities to the fusion protein ( IC50∼0 . 3 µg/ml ) ( Fig . 5C ) . As was observed in the cell-cell fusion assay , the ΔN139INN/K601N virus inhibition curve was shifted to the left , with an IC50 of ∼1 . 3 µg/ml , confirming that ΔN139INN/K601N is more resistant to the CD4-based inhibitor relative to WT and T138N/L494I/K601N in a virion context . The abilities of T138N/L494I/K601N , ΔN139INN/K601N and WT Envs to utilize a panel of CCR5 coreceptor mutants were next compared in order to determine if alterations to the mode of CCR5 engagement could account for reversion . The results presented in Fig . 5D indicate that the T138N/L494I/K601N , ΔN139INN/K601N and WT Envs exhibited almost identical patterns of mutant CCR5 utilization . For example , Q4A and Y14A ( N-terminal domain , Nt ) , H88A ( extracellular loop 1 , ECL1 ) , K171A , E172A and Q188A ( extracellular loop 2 , ECL2 ) , and F264A and R274A ( extracellular loop 3 , ECL3 ) CCR5 mutants supported fusion with the 3 Env constructs to the same extent as WT CCR5 , whereas fusion with Q280A ( extracellular loop 3 ) was decreased to 25–40% of WT CCR5 activity . We also compared the abilities of T138N/L494I/K601N , ΔN139INN/K601N and WT Envs to mediate cell-cell fusion with the CCR5-Y14N tyrosine sulfation mutant , which exhibits a lower affinity for gp120-gp41 and functions as a HIV-1 coreceptor in a cell surface concentration-dependent manner [61] . The coreceptor activity of WT CCR5 for WT and T138N/L494I/K601N remained at consistently high levels for fusion with BHK21 targets cotransfected with a constant amount of pT4luc vector and a dilution series of pc . CCR5 DNA; the fusion activity of ΔN139INN/K601N was slightly diminished across the pc . CCR5 dilution series ( Fig . 5E ) , consistent with the data presented in Fig . 3D . These data are consistent with previous findings indicating that trace amounts of CCR5 are sufficient to mediate efficient fusion and entry by HIV-1 [61] . By contrast , fusion mediated by the 3 Env constructs exhibited a strict dependence on the amount of transfected CCR5-Y14N plasmid without any significant changes in the CCR5-Y14N concentration curves due to the Env mutations being observed . These data indicate that the restoration of function to the mutated DSR by the T138N/L494I or ΔN139INN suppressor mutations in gp120 is unlikely to be a result of altered CD4 and CCR5 utilization . We considered that an increase in the efficiency of a post receptor-binding event , such as 6-helix bundle formation , could aid in the functional compensation of the DSR mutations . We therefore asked if the mutations in gp120 and K601N led to changes in sensitivity to the HR2 synthetic peptide analogue , C34 , which blocks fusion by binding to the coiled coil of HR1 helices in a receptor-triggered prehairpin intermediate conformation of gp41 [17] , [62] . We expect that faster 6-helix bundle folding kinetics would correspond to decreased C34 sensitivity . Figure 5F indicates that T138N/L494I/K601N , ΔN139INN/K601N and WT Envs exhibited almost identical C34 fusion inhibition curves with IC50s of ∼100 nM . These data indicate that the T138N/L494I/K601N reversion mechanism is unlikely to be due to changes in gp120-receptor interactions nor post-receptor binding events such as efficiency of 6-helix bundle formation , whereas ΔN139INN/K601N may be subtly altered in sCD4-induced changes that inhibit membrane fusion function . The DSR mediates association with gp120 and may play a role in the activation of gp41 by responding to receptor-induced changes in gp120 [27] , [28] , [29] , [30] , [31] . To better understand how glycosylation in V1 impacts on gp120-gp41 interactions , we assessed the functional effects of T138N , L494I , T138N/L494I and ΔN139INN mutations on two other gp120 contact residues within the DSR , Leu593 and Trp596 , in addition to Lys601 [27] , [28] . While L593V and K601D mutations resulted in decreased gp120-gp41 association and cell-cell fusion function , the W596L mutant exhibited WT levels of gp120-gp41 association but reduced cell-cell fusion by ∼40% at subsaturating Env ( P<0 . 02 , 2 sample t-test , unequal variances ) ( Fig . 6A , B ) . Figures 6A and B indicate that T138N had no effect on cell-cell fusion when combined with L593V or K601D , whereas small improvements in glycoprotein association and fusion function were observed when L494I was added to the DSR mutants; wild type levels of gp120-gp41 association and fusion were attained when both T138N and L494I were combined with L593V or K601D . By contrast , W596L exhibited WT fusion levels in combination with T138N . The N139INN deletion did not provide any improvement to L593V fusogenicity , but restored increasing levels of fusion function to K601D and W596L , respectively . These data indicate that T138N in V1 alters the gp120-gp41 association site such that fusion function is less dependent on Trp596 and , when L494I is also present , on Leu593 and Lys601 . The N139INN deletion renders fusion function less dependent on Trp596 and Lys601 . Previous studies have shown that larger numbers of potential N-linked glycans in V1V2 correlate with resistance to NAb [34] , [37] , [38] , [39] , [40] , [42] . We therefore asked whether the T138N and ΔN139INN mutations were linked to changes in the neutralization sensitivity of Env-pseudotyped reporter viruses . The results ( Fig . 7 ) show step-wise ∼0 . 5log10-increases in the neutralization sensitivities of T138N and ΔN139INN to the monoclonal NAb 2G12 , which is directed to a glycan cluster involving Asn295 , Asn332 Asn339 , Asn386 and Asn392 on the outer face of gp120 [63] , [64] , [65] , [66] , [67] , [68] ) . The 2G12 IC50s for WT , T138N and ΔN139INN were 13 , 4 and 1 . 5 µg/ml , respectively . In the case of PG16 , which recognizes an epitope in V1V2 involving the Asn156 and Asn160 oligomannose glycans [7] , [69] , [70] , neutralization was enhanced for ΔN139INN only ( IC50 = 0 . 075 and 0 . 015; IC90 = 1 . 6 and 0 . 16 µg/ml , respectively , for WT and ΔN139INN ) . By contrast , neutralization by IgGb12 ( CD4 binding site [71] , [72] ) and 2F5 ( membrane proximal external region of gp41 [73] , [74] , [75] ) was not affected by the V1 mutations . Pooled IgG from HIV-1 infected individuals ( HIVIG ) was used as a reference reagent . In this case , a reproducible ∼2-fold increase in sensitivity to neutralization by HIVIG was observed with the V1 PNGS mutations ( IC50∼550 µg/ml for WT , 280 µg/ml for T138N and ΔN139INN ) indicating that these glycans are likely to modulate neutralization epitopes recognized by human immune sera . These data indicate that the adjacent V1 glycans shown here to modulate the gp120-gp41 association site are linked to neutralization resistance .
The gp120-gp41 complex of HIV-1 is maintained through interactions between C1 and C5 of gp120 with the DSR of gp41 . Receptor engagement elicits a conformational signal that is transmitted through gp120 and sensed by the gp120-contact residues to activate the membrane fusion potential of gp41 . In this study , we forced the evolution of a DSR mutant ( K601D ) with defective gp120-gp41 association , reasoning that the locations of emergent 2nd site mutations that suppress the defect will point to structural elements within gp120-gp41 with functional links to the DSR . The loss of glycans at Asn136 or Asn142 within V1 , the former acting in conjunction with L494I in C5 and the latter with a D601N pseudoreversion , largely suppressed the functional defects , with Env function becoming less dependent on particular gp120-contact residues ( Leu593 , Trp596 and Lys601 ) within the DSR . One mechanism of reversion involved the acquisition of L494I in C5 that led to slightly improved gp120-gp41 association and fusion function with Asp601 , but a WT-like phenotype also required deletion of the glycan at Asn136 in V1 through T138N . Outgrowth by Asn601-containing clones suggested that the negative charge of Asp601 was not favoured in the long term . Cell-cell fusion and viral replication were also restored by a short ΔN139INN deletion in V1 operating in conjunction with a D601N pseudoreversion . In this case Env fusogenicity was maintained despite a more labile gp120-gp41 complex . The ΔN139INN/K601N phenotype was largely recapitulated by S144N/K601N , suggesting that loss of the Asn142 glycan is sufficient to restore fusion function when Asn601 is present in the DSR . Mutations designed to ablate the remaining PNGSs at positions 141 , 156 , 160 and 186 within V1V2 , did not restore function to Asn601-containing Envs , indicating a specific functional linkage between the DSR and the Asn136 and Asn142 glycans in the context of AD8 Env . It should be noted that V1V2 exhibits high degree of diversity in sequence , length and numbers of PNGSs [34] , [35] , [36] , [37] , [38] , [39] , [40] , [41] , [42] , [43] , [44] . It is therefore likely that the functional linkages between particular V1V2 glycans and residues within the gp120-gp41 association site will vary in a strain-dependent manner . The impacts of the V1 glycan changes on the gp120-gp41 association site were probed further by combining T138N , L494I and ΔN139INN with various DSR mutations: L593V and K601D , which disrupt gp120-gp41 association and fusion function , and W596L which inhibits fusion [27] , [28] . The reverting mutations were found to change the dependence of fusion and association on these DSR residues . For example , T138N/L494I rendered subunit association and fusion function largely independent of Leu593 and Lys601 , whereas less dependence on Lys601 was observed with ΔN139INN . In the case of W596L , T138N and ΔN139INN were sufficient to confer WT fusion levels , but this did not involve changes to the WT-like gp120-gp41 association phenotype of W596L . The fusion gains conferred to W596L may involve transduction of a receptor-induced activation signal from gp120 to gp41 through the association site in manner that is less dependent on Trp596 . The earlier finding that W596L blocks sCD4-induced formation of the gp41 prehairpin intermediate indicates that Trp596 can play a role in receptor-triggered gp41 activation [29] . The generation of functional Env complexes with stable gp120-gp41 association following the combination of V1 glycan mutations with various L593V , W596L or K601D DSR mutations , may be mediated by a shifting of the points of intersubunit contact to conserved residues at alternative positions within the synapse [e . g . Gln591 , Gly597 , Thr606 and Trp610 [27] , [28] , [76]] , and/or other regions of gp41 that appear to interact with gp120 , including the fusion-peptide proximal segment [77] , HR1 [78] , [79] , and HR2 [76] . V1V2 comprises a conserved 4-stranded β-sheet minidomain that is stabilized by 2 interstrand disulfides ( Cys126-Cys196 and Cys131-Cys157 ) . The highly variable segments and PNGSs are for the most part contained within the connecting loops ( Fig . 8A ) [7] . A model of glycosylated AD8 V1V2 based on the CAP45 V1V2 structure [7] indicates that 6 N-linked glycans would encompass this minidomain in a hydrophilic shell ( Fig . 8A ) . The most obvious changes due to T138N and ΔN139INN would be the loss of hydrophilic glycan bulk ( ∼1840 Å3 per high-mannose glycan; see [80] ) and , for the latter , a shortening of the V1 loop ( Fig . 8A ) . N-linked glycans have been found to induce local order in the backbone of protein loop regions by destabilizing the unfolded state , the increased entropy of the flexible glycan side chain compensating for the decreased entropy of the protein backbone [81] , [82] , [83] . The loss of the Asn136 and Asn142 glycans may lead to localized disorder and structural change and/or instability within V1 , which in turn affects other structural elements of the gp120-gp41 complex . Cryoelectron tomography indicates that the Asn136 and Asn142 V1 glycans would be located at the apex of the trimeric Env spike whereas the gp120-gp41 association synapse is underneath the gp120 trimer [8] , [12] . An allosteric mechanism whereby changes in V1 can affect the structure of the distal gp120-gp41 association site is suggested by the architecture of gp120 , wherein 3 structurally plastic layers of the core domain are linked to the N- and C-terminal gp41-associating segments via an invariant 7-stranded β-sandwich [8] ( Fig . 8B ) . Layer 2 connects the β2-β3 hairpin that forms the base of V1V2 via the β-sandwich to the gp120 N- and C-terminal segments ( Fig . 8A , B ) . Thus changes in V1V2 structure and/or stability due to the loss of the Asn136 or Asn142 glycans may cause a distortion or displacement of layer 2 , which is translated to the N- and C-terminal gp41-interacting segments via the β-sandwich . Our data suggest that the glycan changes in V1 observed here involve a structural remodeling of the gp120-gp41 association interface such that Env function is less dependent on particular gp120-contact residues within the DSR with conserved residues at alternative positions within the synapse being employed for intersubunit contact [27] , [28] . The V1V2 minidomain has been shown to modulate the accessibility and/or conformation of the CD4-binding site and V3 loop [38] , [84] , [85] , [86] , [87] , the latter of which mediates chemokine receptor binding . Our data indicate that the sCD4-binding abilities of monomeric gp120 molecules bearing 2nd and 3rd site mutations in various combinations were not significantly different to the WT , suggesting that changes in CD4-binding ability per se do not contribute to the mechanism of reversion . However , ΔN139INN/K601N-mediated cell-cell fusion and virus infectivity were more resistant to sCD4 inhibition suggesting that the conformation of the ΔN139INN/K601N Env trimer is subtly different to that of WT and T138N/L494I/K601N . sCD4 has been shown to primarily inhibit HIV-1 infection by inducing a transiently activated glycoprotein complex that rapidly undergoes irreversible conformational changes linked to a loss of function [88] . This sCD4-mediated inactivation correlates with a rapid decay in exposure of the HR1 coiled coil groove of the fusion-activated gp41 prehairpin intermediate . It may be that the apparently looser association between gp120 and gp41 of ΔN139INN/K601N alters the conformational pathway of the sCD4-induced state to enable maintenance of fusion competence . A striking feature of gp120-gp41 is the occlusion of the conserved protein surface by a glycan shield comprising ∼24 N-linked glycans on gp120 and 4–5 on gp41 [3] , [56] , [89] , [90] , [91] . An evolving glycan shield is an important mediator of viral escape from NAbs where an increase in the number and/or a change in the position of PNGSs alter NAb sensitivity [35] . A number of studies have indicated that V1V2 plays a particularly important role in regulating neutralization resistance [36] , [38] , [41] , [42] , which generally correlates with V1V2 elongation and insertion of PNGSs , in some cases at and C-terminal to position 136 ( Fig . 8C ) [34] , [37] , [38] , [39] , [40] , [42] , [43] , [44] . Our studies indicate that subtle changes to the glycan shield in V1 impact on neutralization by glycan-dependent brNAbs . For example , the Asn136 and Asn142 glycan mutations were found to increase the sensitivity of HIV-1 pseudovirions to the 2G12 NAb , which is dependent on high-mannose glycans on the outer face of gp120 , including those attached to Asn295 and Asn332 at the base of V3 , as well as Asn339 , Asn386 and Asn392 [63] , [64] , [65] , [66] , [67] , [68] ( Fig . 8B ) . Neutralization epitopes within V3 appear particularly sensitive to changes in V1V2 , which is likely due to the proximity between these variable structures in the context of trimeric gp120-gp41 [10] , [12] , [38] , [84] , [86] . The Asn136 and Asn142 V1 glycan deletions may modulate the structure and/or accessibility the 2G12 glycan epitope by altering V1V2-V3 interactions in the context of trimeric gp120 . Alternatively , the enhanced neutralization of T138N and ΔN139INN by 2G12 may be a result of changes in the global antigenic structure of gp120-gp41 , as was suggested by the finding that the V1 glycan deletions slightly enhanced the neutralization potency of polyclonal HIVIG . That V1V2 can regulate global antigenic structure via changes in the glycan shield has been suggested by the finding that glycosylation changes within gp120 at 197 , 234 , 295 and 301 contribute to the restoration of infectivity to viruses from which the entire V1V2 region had been deleted [92] . We also observed that ΔN139INN led to increased sensitivity to neutralization by NAb PG16 , whose complex epitope includes the V1V2 glycans at Asn156 and Asn160 and is influenced by residues in V3 [7] , [69] , [70] . The AD8 V1V2 model suggests that the Asn142 glycan is proximal to the Asn156 glycan , implying that the N139INN deletion may relieve a steric constraint that enables better epitope access for PG16 . Alternatively , or additionally , changes in the disposition of V3 , as was implied by the increased sensitivity of ΔN139INN to neutralization by 2G12 , may contribute to the increased neutralization efficacy of PG16 . Overall , our data indicate that the Asn136 and Asn142 glycans of V1 can modulate local ( PG16 ) and remote ( 2G12 ) glycan-dependent neutralization epitopes as well as the global antigenic structure of Env ( HIVIG ) and that these changes are functionally linked to a remodelling of the gp120-gp41 association site . Our data also imply that glycan shield evolution may indirectly affect the inhibitory potential of novel fusion blockers , such as PF-68742 , for which the DSR is a component of their targeting mechanism [33] . Conversely , DSR sequence evolution driven by potential DSR-directed entry inhibitors may be associated with compensatory V1 glycan changes as described here and the coevolution of neutralization sensitivity . Our previous work indicated that the gp120-gp41 association interface is structurally and functionally plastic despite exhibiting a high degree of sequence conservation [28] . In this study , we have found that changes at the 136 and 142 V1 glycans that are associated with neutralization sensitivity appear to remodel the gp120-gp41 association site , rendering certain highly conserved gp120-contact residues ( i . e . Leu593 , Trp596 and Lys601 ) less important for gp120 association and fusion function , and thereby implying that gp120-gp41 contact residues at alternative positions within the synapse become utilized for these functions . The allosteric modulation of the conserved DSR-C1–C5 synapse by distal V1 glycans may represent a mechanism whereby functionally relevant gp120-gp41 association is maintained as the virus acquires neutralization resistance due to the evolution of its glycan shield .
The preparation of the cytomegalovirus promoter-driven HIV-1AD8 Env expression vector , pCDNA3 . 1-AD8env , is described elsewhere [28] . pΔKAD8env was derived by religation of the end-filled HindIII and EcoRI sites of pCDNA3 . 1-AD8env . In vitro mutagenesis of the gp41 region was carried out by overlap extension PCR . Mutants of the pAD8 infectious clone ( obtained from K . Peden [93] were prepared by transferring the EcoRI-BspMI env-containing fragment from pCDNA3 . 1-AD8env vectors into pAD8 . Bacteriophage T7 promoter-driven gp120 expression vectors , based on pTM . 1 [94] , were generated by ligating PCR-amplified HIV-1AD8 gp120 fragments into the NdeI and StuI sites of pTMenv . 2 [95] to give pTM-AD8gp120 . PBMC infections were conducted as described previously [96] . Briefly , PBMCs isolated from buffy packs ( Red Cross Blood Bank , Melbourne ) were stimulated with phytohemagglutinin ( 10 µg/ml; Murex Diagnostics ) for 3 days in RPMI 1640 medium containing 10% fetal calf serum and interleukin-2 ( 10 units/ml; Boehringer-Mannheim ) . Virus stocks were prepared by transfecting 293T cell monolayers with pAD8 infectious clones using Fugene 6 ( Roche ) . Virus-containing transfection supernatants were normalized according to reverse transcriptase ( RT ) activity , and then used to infect 105 PBMCs in a 96-well tissue culture plate ( eight 10-fold serial dilutions of each virus were tested in triplicate ) . The supernatants were assayed for RT activity at various time points . Phytohemagglutinin-stimulated PBMCs were infected with equivalent amounts of wild type ( WT ) and K601D-mutated HIV-1AD8 ( according to RT activity ) in parallel and maintained in culture for 10 days . Cell-free culture supernatants were filtered ( 0 . 45 µm pore size ) and normalized according to RT activity prior to the next passage ( 5 passages in total ) . Genomic DNA was extracted from infected PBMCs using Qiagen DNeasy . The viral DNA fragment encompassed by nucleotides 5954–9096 ( HXB2R numbering convention ) was PCR-amplified using Expand HiFi ( Roche ) and the primers , 5′-GGCTTAGGCATCTCCTATGGCAGGAAGAA ( Env1A ) and 5′-TAGCCCTTCCAGTCCCCCCTTTTCTTTTA ( Env1M ) [97] . The amplified sequences were ligated into pGEM-T or pΔKAD8env ( KpnI-XbaI ) and the entire env open reading frame sequenced using ABI BigDye terminator 3 . 1 . Lysates of Env-expressing 293T cells or virions pelleted from pAD8-transfected 293T cell supernatants were subjected to SDS-PAGE under reducing conditions , transferred to nitrocellulose and then probed with mAb C8 to gp41 ( from G . Lewis [98] , DV-012 to gp120 ( from M . Phelan [99] , [100] , or mAb 183 to CA ( from B . Chesebro and K . Wehrly [101] , [102] ( AIDS Research and Reference Reagent Program , NIAID ) as described [28] . Cell-cell fusion assays were conducted as described [77] . Briefly , 293T effector cells were cotransfected with pCDNA3 . 1-AD8env or pΔKAD8env and pCAG-T7 [103] plasmids , while BHK21 target cells were cotransfected with pT4luc [27] and pc . CCR5 ( AIDS Research and Reference Reagent Program from N . Landau [104] or a panel of CCR5 mutants in the pcDNA3 expression vector ( kind gifts of J . S . Sodroski and R . W . Doms [105] , [106] ) . The Y14N mutation was introduced to pc . CCR5 using the Quikchange II XL kit ( Stratagene ) . At 24 h posttransfection , targets and effectors were cocultured in triplicate in a 96-well plate ( 18 h , 37°C ) and then assayed for luciferase activity ( Promega SteadyGlo , Madison , Wis . ) . The sensitivities of WT and mutant Env proteins to the fusion inhibitor peptide C34 ( WMEWDREINNYTSLIHSLIEESQNQQEKNEQELL; Mimotopes , Australia [62] ) were determined by coculturing effector and target cells in the presence of serially diluted C34 . The sensitivities of WT and mutant Env proteins to sCD4 ( NIH AIDS Research and Reference Reagent Program ) were determined by incubating the Env-expressing 293T cells with a dilution series of sCD4 for 3 . 5 h followed by coculturing the effector and target cells in the presence of sCD4 for 8 h . Single cycle infectivity assays were conducted as described [77] . Briefly , Env-pseudotyped luciferase reporter viruses were produced by cotransfecting 293T cells with pCDNA3 . 1-AD8env or pΔKAD8env vectors plus the luciferase reporter virus vector , pNL4 . 3 . Luc . R−E− ( AIDS Research and Reference Reagent Program , N . Landau [107] ) , using Fugene 6 . The infectivity of pseudotyped viruses was determined in U87 . CD4 . CCR5 cells ( AIDS Research and Reference Reagent Program , H . Deng and D . Littman [108] ) . 293T cells were transfected with pCDNA3 . 1-AD8env or pΔKAD8env vectors . At 24-h posttransfection , the cells were incubated for 30 min in cysteine and methionine-deficient medium ( MP Biomedicals , Seven Hills , NSW , Australia ) , and then labelled for 45 min with 150 µCi Tran-35S-label ( MP Biomedicals ) . The cells were then washed and chased in complete medium for 5–6 h prior to lysis . Cell lysates and clarified culture supernatants were immunoprecipitated with IgG14 or HIVIG and protein G Sepharose and subjected to SDS-PAGE in the presence of ß-mercaptoethanol [28] . The labelled proteins were visualized by scanning in a Fuji phosphorimager . Quantitation of bands was performed using Image Gauge ( FUJIFILM ) software . To determine the effects of the competitive α-mannosidase inhibitor , kifunensine ( Sigma ) , on Env synthesis and processing , cell-cell fusion and gp120-gp41 association , the compound was added to 293T cells at the time of transfection , and was maintained at the concentrations indicated in the results section through each step of the assays described above . 293T cells were cotransfected with pTM-AD8gp120 and pCAG-T7 vectors using Fugene 6 . At 24-h posttransfection , the cells were incubated for 30 min in cysteine and methionine-deficient medium , labelled for 45 min with 150 µCi Tran-35S-label , and then washed and chased in complete medium for 6 h . The clarified culture supernatants were adjusted to 0 . 6 M KCl , 1 mM EDTA and 1% w/v Triton-X100 and the gp120 content quantified following immunoprecipitation with IgG14 and protein G Sepharose , reducing SDS-PAGE and scanning in a Fuji phosphorimager . Equivalent amounts of WT and mutant gp120 proteins were incubated with serial dilutions of sCD4 ( 1 h , room temperature ) and then incubated with mAb OKT4 and BSA-Sepharose on a vibrating platform ( 30 min , room temperature ) . After pelleting the BSA-Sepharose , protein complexes were coprecipitated using protein G-Sepharose and gp120 quantified following reducing SDS-PAGE and scanning in a Fuji phosphorimager . Wild type , T138N/L494I/K601N and ΔN139INN/K601N virus stocks were adjusted such that ∼2×106 relative light units ( RLU ) were obtained following a 48 h-infection of TZM-bl cells , a HeLa cell line expressing CD4 and CCR5 and harbouring integrated copies of the luciferase and β-galactosidase genes under control of the HIV-1 promoter ( obtained from J . C . Kappes , X . Wu and Tranzyme Inc . , NIH AIDS Research and Reference Reagent Program [109] , [110] , [111] ) . A dilution series of CD4-IgG2 , a tetrameric fusion protein comprising human IgG in which the Fv domains have been replaced by domains 1 and 2 of human CD4 ( Progenics Pharamceuticals , NIH AIDS Research and Reference Reagent Program ) , was incubated with the viruses for 4 h at 37°C prior to addition to TZM-bl cells ( 200 µl; 104 cells in 96-well culture plates ) . The cells were lysed 48 later and assayed for luciferase activity ( Promega ) . Neutralization assays were conducted using a modification of the method of Dhillon et al . [112] . Briefly , a solution of Env-pseudotyped luciferase reporter viruses determined previously to give ∼5×105 RLU following infection of U87 . CD4 . CCR5 cells was mixed with an equal volume of serially diluted IgG and incubated for 1 h at 37°C . One hundred microliters of the virus-IgG mixtures was then added to U87 . CD4 . CCR5 cells ( 104 cells per well of a 96-well tissue culture plate , 100 microlitres ) and incubated for 2 days prior to lysis and assay for luciferase activity ( Promega , Madison , WI ) . Neutralizing activities for antibody samples were measured in triplicate and reported as the average percent luciferase activity . Purified IgG of monoclonal NAbs 2F5 , 2G12 and IgGb12 were obtained from Polymun Scientific ( Austria ) , PG16 was obtained from the International AIDS Vaccine Initiative , while HIVIG was obtained from F . Prince through the NIAID AIDS Research and Reference Reagent Program . A homology-based model of HIV-1AD8 V1V2 was generated from PDB entry 3U4E [7] using the Modeller algorithm [113] within Discovery Studio , version 3 . 0 ( Accelrys ) . Five independent models were generated from iterative cycles of conjugate-gradient energy minimization against spatial constraints derived from the template crystal structure . The model with the lowest energy ( probability density function ) was glycosylated in silico with oligomannose side chains using the glycosciences . de server ( http://www . glycosciences . de/modeling/glyprot/php/main . php ) [114] , [115] . | The envelope glycoprotein gp120-gp41 complex of HIV-1 mediates receptor attachment and virus-cell membrane fusion , leading to cellular entry . A shield of asparagine-linked oligosaccharides occludes the gp120-gp41 protein surface and evolution of this glycan shield provides a means for evading circulating neutralizing antibody . Here we examined how conserved structural elements of the glycoprotein complex , in particular the gp120-gp41 association site , retain functionality in the context of glycan shield evolution . This information is important for the evaluation and exploitation of such conserved functional determinants as potential drug and/or vaccine targets . Our data indicate that the loss of either of 2 adjacent glycans in variable region 1 of gp120 leads to changes in local and remote glycan-dependent epitopes and that this is linked to a remodelling of gp120-gp41 interactions in order to maintain a functional gp120-gp41 complex . We propose that this represents a mechanism for the functional adaptation of the gp120-gp41 association site to an evolving glycan shield in a setting of neutralizing antibody selection . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"medicine",
"infectious",
"diseases",
"sexually",
"transmitted",
"diseases",
"medical",
"microbiology",
"virology",
"microbial",
"pathogens",
"hiv",
"retrovirology",
"and",
"hiv",
"immunopathogenesis",
"biology",
"microbiology",
"host-pathogen",
"interaction",
"viral",
"diseases"
] | 2013 | Allosteric Modulation of the HIV-1 gp120-gp41 Association Site by Adjacent gp120 Variable Region 1 (V1) N-Glycans Linked to Neutralization Sensitivity |
Nephron progenitor number determines nephron endowment; a reduced nephron count is linked to the onset of kidney disease . Several transcriptional regulators including Six2 , Wt1 , Osr1 , Sall1 , Eya1 , Pax2 , and Hox11 paralogues are required for specification and/or maintenance of nephron progenitors . However , little is known about the regulatory intersection of these players . Here , we have mapped nephron progenitor-specific transcriptional networks of Six2 , Hoxd11 , Osr1 , and Wt1 . We identified 373 multi-factor associated ‘regulatory hotspots’ around genes closely associated with progenitor programs . To examine their functional significance , we deleted ‘hotspot’ enhancer elements for Six2 and Wnt4 . Removal of the distal enhancer for Six2 leads to a ~40% reduction in Six2 expression . When combined with a Six2 null allele , progeny display a premature depletion of nephron progenitors . Loss of the Wnt4 enhancer led to a significant reduction of Wnt4 expression in renal vesicles and a mildly hypoplastic kidney , a phenotype also enhanced in combination with a Wnt4 null mutation . To explore the regulatory landscape that supports proper target gene expression , we performed CTCF ChIP-seq to identify insulator-boundary regions . One such putative boundary lies between the Six2 and Six3 loci . Evidence for the functional significance of this boundary was obtained by deep sequencing of the radiation-induced Brachyrrhine ( Br ) mutant allele . We identified an inversion of the Six2/Six3 locus around the CTCF-bound boundary , removing Six2 from its distal enhancer regulation , but placed next to Six3 enhancer elements which support ectopic Six2 expression in the lens where Six3 is normally expressed . Six3 is now predicted to fall under control of the Six2 distal enhancer . Consistent with this view , we observed ectopic Six3 in nephron progenitors . 4C-seq supports the model for Six2 distal enhancer interactions in wild-type and Br/+ mouse kidneys . Together , these data expand our view of the regulatory genome and regulatory landscape underpinning mammalian nephrogenesis .
The mammalian metanephric kidney maintains fluid homeostasis . The number of individuals afflicted with kidney disease is on the rise , and reduced nephron number has been associated with disease outcome [1] . In the mouse , genetic studies have demonstrated that nephrons are generated from a Six2+ progenitor pool in a regulatory process requiring the transcriptional action of Six2 for progenitor maintenance [2] . Human SIX2 shows an expression and activity similar to its murine counterpart suggesting that mouse Six2 and human SIX2 likely have similar functions [3] . Consistent with this view , human mutations in SIX2 are associated with renal hypodysplasia and the malignant transformation of progenitor cells in Wilms’ tumor , a pediatric nephroblastoma [4–6] . There is an increasing interest in the relationship between nephron progenitors , their output , and congenital and acquired kidney disease [1 , 7] . Further , new approaches to modulate nephron progenitor outputs to generate kidney structures in vitro call for a better understanding of regulatory processes at play in vivo [8–10] . Nephron progenitor specification and nephron progenitor maintenance are dependent on a number of additional transcriptional regulatory factors including Hoxa/c/d11 , Osr1 , Wt1 , Sall1 , Eya1 , Pax2 , and Six1 . Previous studies of mouse mutants in these genes suggest complex hierarchical interactions amongst these factors [11–27] . Identification of their genomic targets and target regulatory mechanisms are essential to determine the nephrogenic regulatory network . Direct nephron progenitor ChIP-seq studies have identified a broad range of potential transcriptional targets of Six2/SIX2 action in the mouse and human kidney , respectively , and verified predicted enhancer modules for several of these targets [3 , 28 , 29] . Six2 interacts at cis-regulatory modules of genes expressed both in the nephron progenitors and their committed nephron-forming descendants through enhancers co-engaged by differentiation-inducing transcriptional complexes formed in response to canonical Wnt signaling [28 , 29] . Interestingly , a potential role for Hox11 paralogs within Six2-predicted cis-regulatory modules is suggested by the strong enrichment of AT-rich homeobox motifs in Six2 ChIP-seq peaks [28] . The genomic targets of Wt1 have also been analyzed by ChIP experiments of embryonic mouse kidneys [30–32] . Though the approach was not specific to nephron progenitors , these studies revealed the interplay with many genes expressed in , and critical for , nephron progenitors , including Fgf and Bmp family members [30–32] . Sall1 ChIP-seq has also shed light on its active roles in nephron progenitors and repressive actions on development of nascent nephrons , respectively [29] . Interestingly , a subset of Six2- and Sall1-bound regions overlap suggesting these factors co-associate and target analysis predicts genes regulating the nephron progenitor population [29] . With a working model that multi-factor binding will highlight key regulatory nodes of the nephron progenitor pathway [33–37] , we utilized ChIP-seq analysis to identify a subset of putative regulatory elements associated with multiple transcription factors . gRNA/Cas9-mediated ablation of ‘regulatory hotspots’ adjacent to Six2 and Wnt4 highlight the significance of these enhancer elements in regulating target gene expression . Additional analyses of the regulatory landscape surrounding Six2 identified insulator-bound elements which constrain enhancer function . In support of this finding , deep sequencing of the Br mutant mouse identified an inversion of Six2 and Six3 loci altering enhancer specificity . These studies highlight the critical role of multi-factor input and proper enhancer context for directing appropriate target gene expression .
To extend our understanding of the transcriptional regulatory networks operating within mouse nephron progenitors , we developed a transgenic approach to overcome the limited availability and inconsistency of working antibodies for key transcriptional components , and complications that arise from the diverse expression of regulatory factors elsewhere in the kidney . In this transgenic strategy , an epitope-tagged transcription factor-of-interest is expressed exclusively within the nephron progenitor compartment using a Six2 distal enhancer ( DE ) previously shown to recapitulate Six2-like , nephron progenitor restricted expression ( Fig 1A [28] ) . This approach obviates the need to enrich for the progenitor population in whole kidney samples simplifying ChIP procedures and avoiding potential artifacts introduced by tissue dissociation and fluorescence-activated cell sorting ( FACS ) . We also took advantage of established tagging methods which have been utilized to successfully isolate protein:DNA complexes[38–41] . Each transcription factor-of-interest is appended with a BioTag-FLAG ( BF ) epitope at the C-terminus of the target protein . Co-production of an EGFP-BirA enzyme on the transgene through an IRES element also allows both ready visualization of transgenic kidneys and biotinylation of the biotin-recognition motif ( BioTag ) enabling an additional mode of isolation of factor-associated DNA or protein complexes through streptavidin affinity purification ( Fig 1A ) . Though the biotin tagging strategy proved successful ( S1C Fig ) and provided a secondary purification option , we did not utilize it for any ChIP experiments as anti-FLAG antibodies were sufficient for all of the studies presented here . To rigorously assess the efficacy of this strategy and to develop a protocol for whole kidney ChIP , we first generated Six2-BFtg mice to determine whether Six2 ChIP-seq generated with the transgenic line ( Six2-BF ) replicates Six2 ChIP-seq using a Six2-specific antibody ( Six2-ab ) [3] . Six2-BF was restricted to the Six2+ nephron progenitors as indicated by specific detection of the anti-FLAG epitope ( Fig 1C ) . FLAG ChIP-seq from Six2-BF+ kidneys identified 6808 Six2-associated regions in the Six2-BF data , with 90% of these peaks overlapping with Six2-ab peaks ( Fig 1B ) . The two datasets were relatively correlated ( R2 = 0 . 69 ) and , as expected , overlapping peaks were ranked higher than Six2-BF unique peaks indicating the variability in the data reflects marginal peak calls ( S1A Fig ) . The most enriched motif discovered from the top 1000 peaks in the Six2-BF ChIP-seq dataset ( ‘TCANGTTTCA’ , 47% , p-value = 10−1190 , Fig 1D ) matched and agreed with the identified Six2 motif from our previous ChIP studies ( S1B Fig; [3 , 28 , 29] ) . The motif was relatively centered within the peaks suggesting direct binding of Six2 to the motif ( Fig 1D ) . Additionally , electrophoretic mobility shift assays ( EMSA ) utilizing recombinant Six2 and a Six2 motif identified within the Six2-DE showed a strong interaction of Six2 protein with its DNA target ( S2A Fig ) . Mutational analysis on this Six2-binding site demonstrated that the most conserved bases in the consensus ( 1T , 6T and 9C ) were critical individually for effective protein-DNA interaction ( S2A Fig ) . Wt1 and bHLH recognition motifs were also significantly enriched in Six2 binding regions ( S1F Fig ) consistent with an expected role for Wt1 within the progenitor compartment [31 , 42] , and an unidentified role for a bHLH factor . To interrogate the regulatory functions of Six2-BF , we performed GREAT Gene Ontology ( GO ) analysis [43] on Six2-BF peaks . Six2-BF peaks were highly enriched near genes associated with kidney development as reflected by the top GO term ‘ureteric bud development’ ( Fig 1E ) . In summary , the FLAG transgenic strategy robustly reproduced Six2-ab ChIP-seq data generated from wild-type kidneys identifying expected Six2 target and gene associations . These whole kidney-derived datasets significantly extend the depth of Six2 ChIP-seq peaks identified from earlier reports ( [28]: 3907 peaks , [29]: 4306 peaks ) . While our transgenic strategy is useful for targets for which there are no working antibodies or when a progenitor-specific ChIP is desired , expression levels of the tagged protein or affinities of the FLAG antibody versus protein-specific antibodies ( if one exists ) may affect the number of relevant peaks discovered . Interestingly , although peaks identified uniquely with the Six2-ab showed lower levels of enrichment , these peaks still enriched for the Six2 motif at a similar level ( 46% ) and were linked to kidney development GO terms suggesting a biological relevance to the interactions ( S1A Fig ) . As nearly all Six2-BF peaks are contained within the larger Six2-ab dataset ( ~90% , Fig 1B ) , for a more complete analysis of Six2 bound target regions we used the latter dataset for subsequent analyses . Having validated the transgenic strategy for generation of nephron progenitor specific ChIP-seq data , we established additional transgenic mouse lines to identify regulatory interactions mediated by other transcriptional regulators in nephron progenitors . Viable and phenotypically normal founders were generated for Hoxd11 ( Hoxd11-BFtg ) and Osr1 ( Osr1-BFtg ) . Immunostaining with anti-FLAG antibodies confirmed the restriction of Hoxd11-BF and Osr1-BF to Six2+ nephron progenitors and validated the use of both transgenic lines for ChIP-seq analyses ( Fig 1C ) . We also attempted to generate transgenic lines for Wt1 , Hoxa11 , Pax2 , Sall1 , and Eya1 but were unsuccessful in producing any founder animals . Further , we were not able to obtain transgenic progeny which survived past birth from the original Hoxd11-BF founder . These observations suggest transgene and/or transgenic line dependent lethality ( see Discussion ) . To map Hoxd11- and Osr1-associated genomic regions within nephron progenitors , we performed FLAG ChIP-seq on E16 . 5 Hoxd11-BFtg/+ and Osr1-BFtg/+ kidneys identifying 7776 Hoxd11-BF and 5032 Osr1-BF associated regions ( Fig 1D ) . Osr1-BF protein levels were markedly lower and this may account for the lower number of target sites identified ( Fig 1C ) . Both Hoxd11-BF and Osr1-BF peaks showed typical enhancer features: similar to the Six2-BF dataset the majority of the peaks were located >5kb from the transcription start site ( TSS ) within intronic ( Six2-BF:46 . 6% , Hoxd11-BF: 46 . 2% , Osr1-BF: 44 . 7% ) or intergenic regions ( Six2-BF: 46 . 5% , Hoxd11-BF: 48 . 7% , Osr1-BF: 43 . 6% ) ( S1D and S1E Fig ) . GREAT analysis identified an enrichment for both factors near genes associated with processes related to metanephric kidney development ( Fig 1E ) . Using the same workflow adopted above for analysis of Six2 interactions , we identified the top DNA motif enriched in Hoxd11-BF ( ‘TTTATGG’ , 38% , p-value = 10−1033 , Fig 1D ) and Osr1-BF datasets ( ‘GCTNCTG’ , 45% , p-value = 10−1438 , Fig 1D ) . Both motifs were well-centered within each peak dataset ( Fig 1D ) . Multiple Hox factors are expressed in nephron progenitors and each may exhibit distinct binding preferences . While the predicted Hoxd11 motif has a prominent AT-rich Hox factor consensus feature , the motif differs from that identified through protein-DNA binding microarray ( PBM ) studies in vitro ( ‘TTTACGA’ , [44] , S2B Fig ) . EMSA analysis confirmed Hoxd11 binding and the relative importance of the bases 2T and 4A which are conserved in both the PBM and ChIP-seq based predictions , while the 1T and the 5T/C positions , which differed between the two predicted motifs , were not important for binding in vitro ( S2B Fig ) . The Osr1 motif identified from our ChIP-seq data closely resembled that predicted from PBM studies ( ‘GCTACTG’ , [44] ) though no strong preference for the 4th nucleotide position was seen in the in vivo motif . EMSA demonstrated Osr1 bound to the predicted Osr1 binding site within the Six2-DE ( GCTGCTG ) . Interestingly , substituting an A in the 4G position to more closely reflect the PBM motif enhanced the Osr1 interaction ( S2C Fig ) . These findings suggest that in vivo regulatory processes may prefer weaker binding , potentially adding greater flexibility to transcriptional interactions . Wt1 and bHLH motifs were also enriched in each peak dataset , as was observed for Six2-BF peaks ( S1F Fig ) . A Wt1-like binding motif was predicted within all three datasets suggesting Wt1 co-regulation within Six2 , Hoxd11 and Osr1 transcriptional networks . Other groups have published Wt1 ChIP from the whole embryonic kidney or glomerulus [30–32] but no nephron progenitor-specific Wt1 data has been generated . We attempted to generate a viable Wt1-BF transgenic line but failed , so we adopted a recently developed protocol for enriching nephron progenitors by magnetic-activated cell sorting ( MACS ) [45] , and performed ChIP-seq with a Wt1-specific antibody on E16 . 5 nephron progenitors ( Wt1-NP , S3A Fig ) . Compared to a Wt1 ChIP from the whole kidney ( Wt1-kidney ) which we generated from the same stage ( S3A Fig ) , the recovered motif from the Wt1-NP dataset , ‘CCTCCCCCNC’ , closely matches the motif identified in our own whole kidney dataset , and published non-nephron progenitor-restricted Wt1 kidney ChIP data ( S3B Fig , [30–32] ) . The motif also matched the predicted Wt1 motif that was highly enriched in the earlier Six2 , Osr1 , and Hoxd11 datasets ( S1F Fig ) . The motif was centered in the ChIP peak dataset supporting direct DNA binding ( S3B Fig ) . The nephron progenitor-specific Wt1 ChIP shared >50% of peaks with our whole kidney dataset . Shared target genes with roles in kidney development were focused on genes involved nephron progenitor maintenance and differentiation , while those unique to the whole kidney also targeted genes associated with podocytes ( S3F Fig ) . This suggests that our Wt1-NP ChIP is representative of regulatory functions for Wt1 within nephron progenitors . The majority of peaks showed an intergenic ( 35% ) and intronic ( 33% ) distribution ( S3D Fig ) . However , Wt1 showed significant enrichment near promoters within 5kb of the TSS ( 25% , S3C and S3D Fig ) , significantly more promoter enrichment than observed with the other factors ( between 3 . 1 and 5 . 3% , S1E Fig ) , potentially reflecting Wt1’s binding preference to a cytosine-rich motif and GC enrichment at promoters . This result is in contrast to the Wt1 ChIP-seq performed by Motamedi et al . , who found peaks to be enriched more distally [31] . However , if we performed GREAT analysis with the ‘basal plus extension’ parameter which includes larger regulatory domains compared to the more restricted ‘single nearest gene’ parameter which was utilized in all of our analyses , we observe a greater enrichment for Wt1-NP peaks 50-500kb from the TSS ( S3C Fig ) . Importantly , the GREAT parameters recovered ‘ureteric bud development’ and ‘metanephric nephron morphogenesis’ terms which are consistent with Wt1 kidney functions ( S3E Fig ) . To investigate potential co-operative actions of Six2 , Hoxd11 , Osr1 , and Wt1 in nephron progenitors , we analyzed all pairwise overlaps of transcription factor binding sites , and evaluated the statistical significance of such two-factor overlap . Not all genome fractions are accessible to transcription factor binding , and binding of many transcription factors correlates with open chromatin [46] . For simplicity , our statistical analysis is built on the assumption that only open chromatin , identified by utilizing the Assay for Transposase Accessible Chromatin with high-throughput sequencing ( ATAC-seq ) within nephron progenitors ( see Methods for details of the approach and access to data ) , is accessible to any of the DNA binding factors analyzed in the current study . We found that the greatest significance of co-binding is observed between Six2 and Hoxd11 ( -log10p = 320 at all Six2 sites where Hoxd11 is bound and -log10p = 361 at all Hoxd11 sites where Six2 is bound ) , and Six2 and Wt1 ( -log10p = 106 at all Six2 sites where Wt1 is bound and -log10p = 123 at all Wt1 sites where Six2 is bound ) interacting regions . The weakest co-association is between Wt1 and Hoxd11 ( -log10p = 6 for each pairwise association ) , although still significant ( Fig 2A ) . Potential target genes for each factor ( based on GREAT analysis , [43] ) were also subjected to pairwise comparisons . Hoxd11 , Osr1 , and Wt1 share the majority of their target genes with Six2 ( ratio greater than 0 . 60 or 60% , Fig 2A and 2B ) . Hoxd11 shows the greatest overlap with Six2 ( 0 . 80 or 80% ) , although all pairwise overlaps showed that nearly half of the comparators target genes are shared with any one factor . These results suggest that these factors likely cooperate in regulatory actions within nephron progenitors . Next , we overlapped all four datasets to identify sites where all factors converge in the potential regulation of target genes . We recovered 373 putative cis-regulatory modules where Six2 , Hoxd11 , Osr1 , and Wt1 associated within 1kb of each other ( Fig 2B , S1 Table ) . Regions co-bound by all four factors displayed the strongest Six2 binding . In addition , Six2 peaks bound by any three-factor combination were on average stronger than two-factor combinations , while Six2 peaks bound by any factor in combination with Six2 were stronger than Six2-only peaks ( S1H Fig ) . We refer to regions co-bound by all four factors as ‘regulatory hotspots’ hypothesizing that these may play a key role in nephron progenitor programming . Consistent with this view , regulatory hotspots were enriched around genes annotated to developmental processes such as ‘ureteric bud development’ ( Fig 2C ) . Further , two regulatory hotspots are known from published studies to drive transgenic reporters with expression profiles reflecting the putative target genes: a region ~60kb upstream of Six2 which corresponds to the Six2-DE used in our transgenic strategy and the Wnt4-DE 50kb upstream of Wnt4 ( Fig 2D , [28] ) . Sall1 is also bound at the Six2 distal enhancer but not at the Wnt4 enhancer site[29] . Six2 is largely restricted to the nephron progenitors while Wnt4 expression is absent from nephron progenitors but activated on progenitor induction in the formation of differentiating renal vesicles [2 , 47] . Thus , engagement of the four factors can occur on target genes for nephron progenitors or genes activated shortly after the onset of nephrogenesis . Other putative targets of regulatory hotspots include Fgf9 which is expressed by nephron progenitors and is involved in regulating their maintenance [48] , and Pax8 which regulates nephron progenitor differentiation [49] ( S1 Table ) . Tsc22d1 and Mgat5 also represent putative targets and knockouts of these genes are reported to generate kidney phenotypes [50 , 51] ( S1 Table ) . Regulatory information may also converge on a common target through alternative enhancer usage . To examine this possibility , we intersected the predicted target gene sets for each factor and identified 1744 genes sharing Six2 , Hoxd11 , Osr1 , and Wt1 associated peaks ( Fig 2B , S2 Table ) . The set of genes identified as having all four factors co-associated at one putative cis-regulatory module or dispersed through multiple interactions sites are predicted to define a set of genes with a significant role in nephron progenitors or their derivatives; we termed this group ‘core targets’ ( Fig 2C , S2 Table ) . This set includes genes expressed in nephron progenitors and implicated in progenitor maintenance and self-renewal including Six2 , Pax2 , Sall1 , Sox4 , and Gas1 [2 , 19 , 52–54] . However , the ‘core targets’ also included genes normally activated downstream in the induced/developing nephron such as Wnt4 , Lhx1 , Pax8 , Hes1 , and Irx1/2 [47 , 49 , 55–57] . To determine whether interactions amongst these transcription factors exist in vivo , we performed immunoprecipitations with Six2 antibodies from E16 . 5 kidney nuclear lysates . Six2 was able to co-immunoprecipitate Hoxd11 and Wt1 ( Fig 2E ) ; however , the absence of a working Osr1 antibody precluded analysis of this factor although recent studies show Six2 and Osr1 complex in vitro [17] . Six2 is also purported to complex with Sall1 [29] though we could not replicate this interaction with available antibodies in our assay . Taken together , these data provide evidence for endogenous , multi-protein complexes among three of the four factors . We sought to identify whether Six2 , Hoxd11 , Osr1 , and Wt1 are each involved in activating or repressing gene expression in nephron progenitors . First , we generated RNA-seq expression profiles of E16 . 5 Six2TGCtg/+ kidney cortex preparations FAC-sorted for GFP+ ( Six2+ ) or GFP- ( Six2- ) cells . Six2+ cells would represent the nephron progenitor population ( both self-renewing and recently induced ) and Six2- cells would largely represent stromal cells as well as ureteric bud tip cells and endothelial cells . Genes with a TPM ( Transcripts Per Kilobase Million ) value >5 and a fold difference >3 between the two cell types were identified: 246 genes were enriched in the Six2+ fraction and 545 genes were enriched in the Six2- cortex fraction ( Fig 3A , S3 Table ) . We asked whether ChIP-seq peaks of any of the transcription factors or the regulatory hotspots are preferentially located adjacent to differentially expressed genes . The results show that peaks from all ChIP-seq datasets occur significantly more often around genes enriched in the Six2+ cells ( Fig 3C ) consistent with a specific role in regulating the nephron progenitor cell versus other cell types of the kidney cortex . However , regulatory hotspots near Foxd1 , a marker of self-renewing stromal progenitors [58] , and Wnt11 , a ureteric tip marker required for normal kidney development [59] ( S1 Table ) , raises the possibility that the four factors may also work together to repress these genes within nephron progenitors . Nephron progenitor cells can be divided into Cited1+/Six2+ self-renewing progenitors and Cited1-/Six2+ differentiating progenitor cells [60] . To address the relationship between regulatory hotspots and programs of progenitor maintenance or commitment , we performed RNA-seq analysis to identify progenitor-specific and early induction gene sets . For the former , a transcriptional profile was generated for E16 . 5 Cited1+; RFP+ cells from Cited1-nuc-TagRFP-Ttg/+ kidneys while Six2+; GFP+ cells from Six2TGCtg/+ P2 kidneys were used to generate the latter dataset ( S4 Table , [61] ) . As expected , Cited1 levels were appreciably lower in the P2 Six2+ cells ( 200 . 9 TPM in E16 . 5 Cited1+ cells vs . 3 . 8 TPM in P2 Six2+ cells ) while Wnt4 transcripts were markedly increased ( 9 . 0 TPM in E16 . 5 Cited1+ cells vs . 219 . 1 TPM in P2 Six2+ cells ) supporting our classification of these datasets ( S4 Table ) . As expected , a comparison of the genes with a TPM >5 and a fold difference >3 between the two cell types showed self-renewing nephron progenitor-specific genes such as Cited1 and Osr1 enriched in the E16 . 5 Cited1+ cell dataset whereas genes involved in progenitor differentiation such as Pax8 and Wnt4 were enriched in the P2 Six2+ cell dataset ( Fig 3B , S4 Table ) . Six2 and Hoxd11 displayed similar enrichment near genes up-regulated in either self-renewing nephron progenitors or in differentiating progenitors ( 1 . 4–1 . 5 fold; Fig 3C ) consistent with roles in promoting the progenitor state , and either preventing or priming nephron forming programs . Osr1 and Wt1 interactions were slightly enriched near genes associated with self-renewing nephron progenitors ( 1 . 4-fold vs 1 . 0-fold for Osr1 , 1 . 2-fold vs 1 . 0-fold for Wt1; Fig 3C ) . Interestingly , the regulatory hotspot associated gene lists showed a higher enrichment around genes upregulated in differentiating cells versus self-renewing progenitors ( 1 . 6-fold vs . 1 . 3-fold; Fig 3C ) . Next , for each single factor or combination of factors we compared the percent of target genes in distinct transcriptional categories: nephron progenitor enriched ( E16 . 5 Six2+ cells ) , self-renewing nephron progenitor enriched ( E16 . 5 Cited1+ cells ) , or differentiating nephron progenitor enriched ( P2 Six2+ cells ) relative to the whole transcriptome . Target genes unique to any single factor were not enriched in any of these categories ( ≤1 . 6% for each ) compared to the whole transcriptome ( ≤1 . 6% for each ) suggesting that single factor input has no particular relevance to nephron progenitor function . Similar observations hold when Hoxd11 co-targeting is examined with Osr1 and Wt1 , ( ≤1 . 8% ) . but not with a Six2 binary combination ( ≥2 . 5% ) suggesting that Hoxd11 has a strong preference for co-regulation of target genes with Six2 ( Fig 3D ) . Generally , the greatest enrichments are observed when all four factors are bound near the target gene in any category ( 2 . 4–7 . 2% ) consistent with co-regulatory input by multiple factors impacting target gene regulation to the greatest extent . In agreement with our earlier analyses , the four-factor overlap has a preference for genes expressed upon differentiation rather than in self-renewing progenitors ( 5 . 9% compared to 2 . 4%; Fig 3D ) . While we have described target genes with known functions in kidney development , we wanted to identify potentially novel candidate genes which are targets of co-regulation either by one cis-regulatory module or dispersed through multiple interactions sites . Target genes of interest include Shisa2 and Shisa3 which are enriched in self-renewing nephron progenitors ( S2 Table ) . Shisa2 is a modulator of Wnt and Fgf signaling , specifically attenuating such signals . The majority of mutant mice exhibit dwarfism and half die postnatally . Shisa3 is a related family member although no overt phenotype was observed for the null allele [62] . Pdgfc and Pdgfa are enriched in the self-renewing and differentiating progenitors , respectively ( S2 Table ) . Their conserved expression in these cell populations of developing mouse and human kidneys have been reported [63–65] . Pdgfa and Pdgfc double mutants have a reported deficiency in cortical renal mesenchyme , however , the mutant kidney phenotype was not analyzed in detail [66] . Ccnd1 ( cyclin D1 ) is a putative target that shows a nearly 7-fold increase in expression in P2 Six2+ cells versus E16 . 5 Cited1+ cells ( S2 Table ) . In situ hybridization confirms strong Ccnd1 in E15 . 5 pretubular aggregates and early differentiating nephrons ( www . gudmap . org , [67 , 68] ) . This suggests that the regulatory networks may directly modulate cell cycle dynamics and balance progenitor proliferation or alternatively may prime putative enhancers of Ccnd1 for rapid activation upon nephron progenitor induction . Sema5a and Epha4 are predicted targets with a similar ~6-7-fold increase in expression in differentiating progenitors confirmed by in situ studies ( S2 Table; www . gudmap . org , [67 , 68] ) suggesting factor regulation of targets genes controlling local cell interactions . To examine the functional significance of ‘regulatory hotspots” , we focused on Six2-DE ( chr17: 85747271–85749534; Fig 4A ) and Wnt-4 DE ( chr4:137216986–137217756; Fig 5A ) elements previously verified in transgenic reporter assays [28] . To examine the requirement for each enhancer , we used CRISPR/Cas9 gene editing technology to delete each enhancer in B6SJLF1/J mice . The Six2-DE deletion and Wnt4-DE deletion were confirmed in founder lines by PCR and Sanger sequencing of products ( Six2∆DE: chr17:85747284–85749542; Wnt4∆DE: chr4:137216991–137217771 ) . For the Six2-DE knockout , we examined kidneys at E16 . 5 and observed no obvious difference in the size of wildtype , Six2∆DE/+ and Six2∆DE/∆DE kidneys ( Fig 4B ) . Six2+ and Wt1+ nephron progenitors were present in Six2∆DE/∆DE kidneys though Six2 levels appear reduced relative to wild-type embryos ( Fig 4C and 4D ) . Nephron structures were formed as reflected by the presence of podocytes and proximal tubules , labeled by Wt1 and LTL ( Lotus tetragonolobus lectin ) , respectively ( Fig 4C ) . The Six2∆DE/∆DE mice were viable; no phenotype was observed . To more accurately assess the effect of the distal enhancer deletion on Six2 expression , we used qPCR to measure relative Six2 levels in nephron progenitors of E16 . 5 kidneys . A 40% reduction of Six2 mRNA was measured in Six2∆DE/∆DE nephron progenitors compared to wildtype ( Fig 4E , p-value = 0 . 006 ) ; higher levels than in mice heterozygous for a Six2 null allele ( Six2CE/+ , [69] ) where Six2 transcripts were reduced approximately 50% relative to wild-type as expected ( Fig 4E ) . The levels of Pax2 mRNA , which is not dependent on Six2 [2] , were relatively similar across all genotypes showing a Six2-specificity for the Six2-DE deletion . Strikingly , when Six2 levels were further reduced by combining a Six2∆DE allele with a Six2 null allele ( either Six2CE/+ or Six2GCE/+ [69] ) , the resultant Six2∆DE/CE embryos exhibited severely hypoplastic kidneys at E16 . 5 , with a complete absence of Six2+ nephron progenitors , mirroring the phenotype of complete removal of Six2 activity ( Fig 4B and 4D , [2] ) where only a few glomeruli ( Wt1+ ) and tubules ( LTL+ ) have formed by E18 . 5 ( S4 Fig ) . As early as E11 . 5 , at the outset of active kidney morphogenesis , Six2∆DE/GCE kidneys were devoid of Six2+ nephron progenitor cells but filled with Pax8+ differentiating nephron progenitors as in Six2 protein null mutant kidneys ( Fig 4G , [2] ) . Taken together these results demonstrate that Six2-DE accounted for approximately 40% of Six2 expression and by combining one Six2-DE allele with a Six2 null allele , the remaining Six2 mRNA levels ( predicted to be 30% of wild-type levels ) were insufficient for Six2-mediated maintenance of the nephron progenitor state . Next , we investigated a ‘regulatory hotspot’ predicted to function in progenitor differentiation . Deletion of the Wnt4 distal enhancer resulted in mutant kidneys that are ~25% smaller than those from wildtype animals ( p = 0 . 2e-4; Fig 5B and 5C ) . Nephrons developed in Wnt4∆DE/∆DE kidneys as reflected by presence of both LTL+ proximal tubules and Wt1+ podocytes ( Fig 5D ) and Wnt4∆DE/∆DE mice are viable . Interestingly , in situ hybridization revealed that expression of Wnt4 is significantly reduced in renal vesicles but remained largely unchanged in the renal medulla of Wnt4∆DE/∆DE kidneys consistent with an overall reduction of Wnt4 mRNA levels in Wnt4∆DE/∆DE kidneys measured by qPCR ( Fig 5F ) . Thus , the Wnt4-DE plays a functional role in regulating Wnt4 mRNA levels in forming nephrons ( Fig 5E ) . The Wnt4∆DE/∆DE phenotype was less severe than Wnt4 protein null mutants where the severely hypoplastic kidney lacks nephron tubules and glomeruli [47]; indeed , low levels of Wnt4 RNA were detected in Wnt4∆DE/∆DE kidneys ( Fig 5D; arrows in Fig 5E ) . When the Wnt4∆DE allele was combined with a Wnt4GCE protein null allele [69] , Wnt4∆DE/GCE kidney size and nephron structures were further reduced , though kidneys were still larger than Wnt4 null kidneys ( Fig 5B–5D ) and Wnt4 mRNA levels were markedly reduced in whole kidney PCR ( Fig 5E and 5F ) . Taken together these results indicate a dose-dependent reduction in kidney size through reduced nephrogenesis upon decreasing Wnt4 activity . Further , residual levels of Wnt4 activity in Wnt4∆DE/GCE kidneys were sufficient to drive low levels of nephrogenesis . Clearly , the Wnt4-DE plays a role in maintaining appropriate levels of Wnt4 transcripts in the nephrogenic program to ensure a normal program of kidney development . Six2 lies ~60 kb from a related family member Six3 , although significant Six3 expression is not observed in the self-renewing nephron progenitors ( S4 Table ) indicating a specificity in Six2-DE interactions . Topologically associating domain ( TAD ) boundaries have been described as CTCF-enriched sites which serve as insulators and prevent promiscuous enhancer activity [70] . We performed CTCF ChIP-seq using E16 . 5 purified nephron progenitors to identify CTCF-bound regions of the genome ( Fig 6B and S3G and S3H Fig ) . We identified strongly bound CTCF sites between the Six2 and Six3 locus , most of which are consistent with ENCODE data analyzing whole P0 kidney samples ( S5B Fig , ENCODE experiment ENCSR143WOK , submitted by Richard Myers , HAIB , [71] ) and predictions from Hi-C on mouse ES cells ( TAD: chr17:85640660–85680660 , S5A Fig , [70] ) . We hypothesize that this region serves as a TAD boundary to prevent the Six2-DE from engaging the Six3 promoter . Consistent with this view , there is a marked bias in the engagement of regulatory factors in nephron progenitors to the Six2 side of this putative boundary , 5’ to the Six2 transcriptional start site ( Fig 6B ) . With these insights into regulation of Six2 , our attention was drawn to the Brachyrrhine ( Br ) mouse , an X-irradiation induced mutant that displays kidney hypoplasia and frontonasal dysplasia , and maps to the Six2 region of chromosome 17 [72] . Though Br mutants have significantly reduced Six2 expression in the kidney and craniofacial tissues , no mutation has been found in the Six2 transcription unit or within 1 . 8 kb upstream of the start codon which includes the Six2-PE elements [72] . Interestingly , Six2 is ectopically expressed in the developing lens of Br heterozygous and homozygous animals , a normal site of Six3 expression [72] . Given that irradiation induces large-scale genomic rearrangements , we speculated that the Br mutation led to a chromosomal rearrangement that removed Six2 from enhancer ( s ) directing normal regulatory input to the nephron progenitor population , placing Six2 under the control of Six3 regulatory elements normally inaccessible the other side of a CTCF-dependent boundary element . Next generation sequencing and sequence alignment identified the underlying sequence change in the Br mutant ( S1 Supplemental Material and Methods ) . The main feature was a large inversion of 324 , 596bp including both the Six2 and Six3 loci . The inversion moves Six3 ~206kb from the Six2-DE , actually further than in the wild-type organization , but importantly the inversion removes the intervening TAD boundary ( Fig 6B ) . In contrast , Six2 is repositioned on the other side of this boundary element within Six3’s unchartered regulatory territory ( Fig 6B ) . In addition to the inversion , two small deletions were detected: a 4 , 630bp deletion ( chr17:85414584–85419213 ) 5’ to the Six3 TSS in the intron of Camkmt , and a 5bp deletion between the Six2-PE and Six2-DE ( chr17:85743809–85743813 , Fig 6A and 6B ) . The results from the sequencing and computational analysis were confirmed by allele-specific diagnostic PCR assays ( S5C and S5D Fig ) . The inversion also separates the last 4 exons of Camkmt from the rest of the transcription unit . However , Camkmt has a TPM of only 2 . 26 in the E16 . 5 Cited1+ nephron progenitor cells and homozygous mutant mice are viable ( International Mouse Phenotyping Consortium , http://www . mousephenotype . org/ , Release 5 . 0 [73] ) , so Camkmt is unlikely to contribute to the kidney phenotype . The rearrangement predicts: i ) ectopic Six2 expression in Six3’s normal expression domain , the lens , as Six3 enhancers can now target Six2 , and ii ) an abnormal interaction between the Six2-DE and Six3 promoter resulting in ectopic Six3 expression in nephron progenitors . To directly examine interaction of Six2-DE with Six2 and Six3 promoters , we performed 4C-seq [74] using Six2-DE as the view point . As expected , in wildtype kidneys Six2-DE interacts with a broad region that includes Six2 transcription start site ( TSS ) ( Fig 6B ) , with the local maxima 7 . 2 kb upstream of Six2 TSS . Noticeably , Six2-DE interaction was restricted by the TAD boundary between Six2 and Six3 ( Fig 6B and S5B Fig ) and no interaction was observed around the Six3 TSS . In kidneys from Br/+ embryos , strong Six2-DE contacts were now observed in the segment of the inverted region that was repositioned between the Six2-DE and CTCF-bound TAD boundary element ( Fig 6B and S5B Fig ) . As expected , with the loss of one wildtype allele in Br/+ embryos , Six2-DE interactions with Six2 TSS , and in general with the region on Six2 side of the TAD boundary , were significantly reduced . A relatively strong , de novo interaction of the Six2-DE was observed ~15 kb upstream of Six3 TSS ( Fig 6B and S5B Fig ) consistent with the model of Six3 expression driven , at least in part , by the Six2-DE in the Br allele . Importantly , the predicted Six2-DE/Six3 upstream contact in the Br allele occurs over a distance of 200 kb from the Six2-DE to the Six3 TSS , a longer interval than the ~130 kb that separates these non-interacting elements in the wild-type allele ( S5B Fig ) . Therefore , the differential interaction of Six2-DE with Six2 TSS and Six3 TSS between wildtype and Br/+ cannot be attributed to shortened distance from Six2-DE to Six3 TSS . Rather , this data supports specific regulation by Six2-DE to Six2 and Six3 that is defined by the TAD boundary . As a result of the altered chromatin architecture introduce by the genomic inversion , ectopic Six2 expression has been reported in the lens of Br/Br mutants [72] , and Six3 expression was reduced in this structure ( S5E Fig ) . Quantitative PCR detected Six3 expression in the kidneys of Br/+ mice at E13 . 5 ( Fig 6C ) and Six3 protein was detected in Six2+ nephron progenitors ( Fig 6D ) . Br/Br mutants resemble Six2 null mutants and have no nephron progenitors at this stage [2] ( Fig 6C and 6D ) . When Br/Br mutants were examined at E11 . 5 , they showed a similar loss and premature differentiation of nephron progenitors as in Six2 null mutants but interestingly low-levels of Six3 were detected in differentiating progenitors ( Fig 6E ) . Ectopic Six3 was also observed in the cranial base of Br mutants at E14 . 5 concomitant with decreased Six2 levels ( S5F Fig ) . Taken together these data lend additional weight to the importance of the Six2-DE in directing Six2 expression and reveal higher order principles of topological organization acting in conjunction with this enhancer to provide target gene specificity to the regulatory landscape .
Our ChIP studies reveal a complex regulatory architecture of the nephron progenitors . Examining co-binding of the four factors suggests each of these genes is itself a target of their combined actions through auto and cross-regulatory inputs , as are a number of other transcriptional regulatory components important for kidney development and nephron progenitor maintenance such as Sall1 and Pax2 ( S2 Table ) . By combining our data with insight from previous studies , a hierarchical network starts to emerge . For example , mutational analyses have demonstrated a requirement for the Hox11 paralogues to activate Six2 expression in metanephric mesenchyme [12] . Hox11 members complex with Pax2 and Eya1 binding to an enhancer that lies within ~1kb of the Six2 TSS , in the Six2-PE [75] . Hox11 acts as an activator of Six2 activity and mutations in Hox motifs results in loss of reporter activity in transgenic assays [76] . We have also shown that the Hox motif within the Six2-DE is required for reporter activity [28] . Consistent with this data , Hoxd11 is bound at the Six2-DE ( Fig 2D ) . However , we did not observe a significant Hoxd11 association to the Six2-PE as reported [76] . This discrepancy may result from preferential enhancer usage at different developmental stages . Two previous studies assayed reporter activity of the ~1kb Six2-PE at E11 . 5 [75 , 76] while our studies assayed Six2-DE activity at E15 . 5 [28] . Hox11 may be required at the Six2-PE to help initiate Six2 expression , but maintenance of expression may then rely , at least partially , with the Six2-DE where Hoxd11 is engaged at E15 . 5 . Additionally , Osr1 and Wt1 are enriched at the Six2-DE compared to the PE ( Fig 2D ) , as is Sall1 [29] , supporting multifactor input at the DE as an important mechanism of Six2 regulation . However , Six2 is bound at both the PE and DE , though PE association is weaker ( Fig 2D ) , suggesting both may contribute at some level to the maintenance and autoregulation by Six2 itself . Unfortunately , technical difficulties preclude detailed temporal analysis of engagement in the small numbers of cells that are the foundation of the nephron progenitor pool . When assessing the targets unique to any transcription factor combination , the greatest enrichment for genes with expression within the nephron progenitors , either in self-renewing or differentiating cells , generally occurred when they were complexed with Six2 ( Fig 3D ) . Hoxd11 showed the lowest levels of enrichment for these targets when engaged with Osr1 or Wt1 in the absence of Six2 , suggesting that its primary regulatory functions rely on engaging with Six2 . Taken together , these data suggest Six2 acts as a master regulator: co-engagement with Six2 predicts a higher probability of regulatory functions within nephron progenitors . Osr1 has been described as a transcriptional repressor in vertebrate kidney development [77] . Xu et al . showed that Osr1 works with the Groucho family members and represses activation of a Wnt4 enhancer specifically in Six2+ nephron progenitors [17] . Consistent with this result , Osr1 associates with the Wnt4 enhancer in our ChIP assay ( Fig 2D ) . Additionally , other genes that are not present in the nephron progenitors but rather in differentiating structures such as Pax8 and Lhx1 are also bound by Osr1 suggestive of a repressive role ( Fig 3C , S2 Table , [49 , 56] ) . However , Osr1 is also bound near genes actively expressed in nephron progenitors such as Six2 and Osr1 itself ( Fig 3C , S2 Table , [2 , 16 , 69] ) . Therefore , our data suggest a more complex relationship than Osr1 simply repressing transcription at all engaged targets . Further , our previous ChIP studies supported dual roles for Six2 in activating transcription within nephron progenitors but also engaging at targets silent in progenitors but activated as progenitors differentiate towards nephrons [3 , 28] . Similarly , Hox11 has been characterized as an activator , specifically of Six2 expression [76] . Consistent with this view , Hoxd11 , is bound near nephron progenitor-specific genes but like Six2 binding is also prominent around differentiation targets ( Fig 3C , S3 Table ) . Similarly , these observations extend to Wt1 nephron progenitor targets . Engagement most likely reflects dual activator and repressor actions of these complexes and which activity could be dependent on currently unidentified co-bound factors . Conversely , factor engagement at differentiation-specific gene targets may facilitate or enable subsequent activation of enhancers for differentiation-associated genes following the induction of nephron progenitors . In this scenario , multi-factor engagement may be necessary but not sufficient for target activation for differentiation associated genes . Additional factors or modification of existing transcriptional components following progenitor commitment may modify the action of these regulatory complexes . We observed a highly significant overlap of transcription factor binding in nephron progenitors . Motif analysis of each ChIP dataset showed de novo , factor-specific motifs were the most enriched ( Fig 1D ) , supporting direct protein-DNA binding . Additionally , our EMSA assays confirmed factor binding to each motif ( S2 Fig ) . On the other hand , previous studies have shown that Six2 can complex with a number of transcription factors , including Hoxa11 [28] and Osr1 [17] in vitro , and Eya1 and Sall1 both in vitro and in vivo [29 , 78 , 79] . Osr1 has also been shown to interact with Wt1 in vitro [80] . These studies support protein-protein interactions amongst these factors and may account , in part , for the multi-factor co-localization on specific genomic targets . Additionally , our kidney immunoprecipitation data suggests that Six2 can interact with Wt1 and Hoxd11 ( Fig 2E ) , confirming such complexes exist in vivo . However , without confirming the co-association of these factors on any genomic loci at the same time and in the same cell , we can only suggest their combined function . The association of each factor with its own DNA target and co-association with each other adds to the difficulty of predicting the actions of the regulatory circuit . Further , it is likely that there are significant components yet to be discovered . For example , all of the ChIP datasets recovered a bHLH motif amongst the most-significantly enriched motifs ( S1F Fig ) . Whereas Myc is a bHLH transcription factor that has been shown to complex with Eya1 and Six2 in the kidney [79] , and loss of function Myc mutants argue for a role in kidney development [81] , the recovered motif is distinct from the conventional Myc-Max target site [44] , suggesting a role for another , unidentified family member . In addition to identifying target genes with known function during kidney development , we also uncovered novel putative targets of the four factors ( see S2 Table for list of all target genes ) . Bmper is a secreted protein that interacts with Bmp proteins and inhibits their function [82] . Inactivation of Bmper in the kidney leads to mild hypoplasia [83]; Bmp signaling plays important roles in the progenitor self-renewal and differentiation [84] . Six2 and the other factors may help fine-tune the level of Bmp signaling through activation of Bmper . Rspo1 is a secreted protein that binds to G protein-coupled receptors that activate Wnt signaling and its function has been implicated in multiple developmental systems [85] . Rspo1 could have a role in modulating Wnt signaling in the nephron progenitor niche although Rspo1 mutants have no obvious kidney phenotype , these mutants have not been analyzed in depth [85] . We also identified other modulators of Wnt signaling within our data . Shisa2 is reported to attenuate Wnt and Fgf signaling during development [62] . Shisa2 is expressed in the nephron progenitors along with its related family member Shisa3 ( S2 Table ) . Other targets like Tsc22d1 and Mgat5 are reported to display kidney phenotypes . Mgat5 is expressed in differentiating structures including podocytes ( S2 Table , Eurexpress , www . eurexpress . org , [86] ) and shows a glomerular phenotype [51] . Tsc22d1 is expressed in the nephron progenitors ( S2 Table ) and mutants have small kidneys [50] . Given the current associations of known targets with kidney development and disease , it is likely that functional analysis of new targets predicted here will identify additional regulators of mammalian kidney development . From our analyses , the majority of significant targets fall under the control of all four factors . These genes fall into multiple functional categories from transcriptional regulators like Six2 , Sall1 , and Pax2 to signaling factors like Fgf9 and Wnt4 to cell cycle regulators such as Ccnd1 and matrix proteins such as Lamb1 ( S2 Table ) . This suggests that these transcription factors control many different aspects of progenitor cell biology . Fewer targets with known kidney functions emerge from the interaction maps where one of more the factors was not bound at the putative regulatory region ( S5 Table ) . However , Eya1 , Wt1 , and Bcam lacked an Osr1 association in combined factor interaction analysis ( S6 Table ) but are well known for their early roles in the kidney program [18 , 21] . Bcam , encodes a surface receptor which binds laminin and is expressed at increasing levels in differentiating progenitors ( S6 Table ) . Knockouts display glomerular abnormalities suggesting important functions in the kidney [87] . Phgdh , a Six2-independent target with highest expression in nephron progenitors ( S6 Table ) participates in L-serine synthesis and knockouts are embryonic lethal [88] . Enhancers directing Six2-like and Wnt4-like reporter gene expression [28] , identified as ‘regulatory hotspots’ co-bound by Six2 , Hoxd11 , Osr1 , and Wt1 in the data here , were shown to play roles in regulating activity of both gene targets . Kidney phenotypes were observed in embryos homozygous for the enhancer deletion ( Wnt4-DE ) or when combined with protein null mutations ( Six2-DE and Wnt4-DE ) . While the study identified functional enhancer regions , neither works alone in regulating normal transcript levels in the target cell type . An alternative proximal enhancer has been documented for Six2 [76 , 89] . This proximal enhancer lies a few hundred base pairs upstream of Six2’s transcriptional start site and strongly binds Six2 , but not the other regulatory factors analyzed here . Alternative enhancers have not been functionally demonstrated for Wnt4 . In summary , our studies provide evidence to support a focus on multifactor input to prioritize functional analysis of large datasets emerging from ChIP-seq studies . CRISPR/Cas9 deletion of an enhancer region >100kb from the TSS for Sox2 that is co-bound by multiple transcription factors regulating pluripotency ( Oct4 , Sox2 , Nanog , and Klf4 [90 , 91] ) provides another example of this strategy to identify strong , bone-fide components of the regulatory genome . Individual enhancer action depends on the larger context of the chromosomal landscape . Our demonstration that the inversion in Br mutant strain , repositions Six2 and Six3 in a new regulatory landscape modifying enhancer interactions that likely contribute to altered features of each gene’s regulation . Each gene exists in a distinct TAD that is likely enforced by the action of a CTCF-dependent boundary element between the two genes . In Br heterozygous and mutant alleles , Six3 is ectopically expressed in nephron progenitors: the boundary element no longer separates the Six3 promoter from the Six2-DE . We hypothesize that this enhancer , and potentially undefined regulatory information 5’ to this enhancer , dominate over other regulatory information that might be present within the Six3 flanking region . As a result , the Six2-DE drives Six3 expression in nephron progenitor cells while Six3 expression is lost from its normal lens expression domain . Even though Six3 was detected in nephron progenitors in Br/+ mutants , Six3 is a member of a functionally divergent sub-group of Six factors [92] . Consequently , Six3 activity failed to compensate for loss of Six2 and Br/Br mutants resemble Six2 null mutants [72] . Interestingly , even though there is no alteration in Six2-PE position relative the Six2 gene , the Six2-PE is not sufficient to drive levels of Six2 which maintain nephron progenitor development in the context of the inversion . Thus , if the Six2-PE were capable of sustaining normal Six2 levels , the inversion may prevent Six2-PE engagement with regulatory factors necessary for its activation . Alternatively , there may be distinct enhancers other than the Six2-PE that are required for Six2’s expression . Six2-bound putative regulatory regions lie upstream of Six2-DE ( S8 Table , Fig 6B ) and these would be predicted to disengage from Six2 regulation in the Br inversion . Topological domains are highly conserved between cell types and across mammalian species [70] . Recent studies have shown that alterations in TADs and CTCF site orientation can affect chromosome architecture and result in altered gene expression [93 , 94] . Specifically , several limb malformations in the human were attributed to the rearrangement of TADs and disrupted boundaries . When genetically modeled in the mouse , altered gene expression suggests a mechanism for driving the limb malformations [93] . The type of topological rearrangements described here could play a role in a subset of the congenital anomalies of the kidney and urinary tract ( CAKUT ) syndrome . Importantly , these micro-rearrangements would not be detected in traditional exome screens . Even whole genome sequencing approaches required tailored alignment algorithms to uncover the junction fragments for the rearrangements as performed here . Together , these studies highlight the importance of non-coding DNA and chromatin architecture to the appropriate regulation of gene expression and the resulting phenotypic consequences incurred by rearranging the regulatory landscape . In addition to Six2 , Hoxd11 , and Osr1 , we attempted to generate transgenic lines for other important regulatory factors including Wt1 , Hoxa11 , Pax2 , Sall1 , and Eya1 . Our goal was to build an extensive regulatory network for the nephron progenitor population and more precisely identify the targets and combinatorial actions of these major players in vivo . However , despite considerable efforts , we were unable to establish correctly expressing founder lines for these factors . The Six2-DE is not only active in the kidney but Six2 is expressed in the developing brain , ear , tendons , and smooth muscle [95] and we observed transgene expression in the brain and ear . Some transgenic lines showed a circling behavior , consistent with inner ear defects , along with insufficient kidney expression and thus were not utilized . The ectopic action of a sub-set of factors in these other sites of Six2-DE activity may have resulted in severe defects and subsequent lethality . Alternatively , there could be dominant effects within the kidney itself from elevating levels of that factor in the normal nephron progenitor context , though this seems less likely given the absence of a kidney phenotype in Six2 , Hoxd11 , or Osr1 transgenic strains , a Six2-BF binding profile that was comparable to the native Six2 protein , and the levels of Six2-DE activity . Despite these technical limitations , we were able to generate a core transcriptional network of four factors important for kidney development . Overlap with additional factors may not add much greater insight; comparison with Sall1 and β-catenin targets reveals many of the same nephron progenitor-specific target genes and a lack of nephron progenitor relevant independent regulation by these factors ( S1 , S5 and S6 Tables ) . Therefore , the data presented here is likely to highlight some of the most critical regulatory elements and target genes which modulate nephron progenitor programs .
All surgical procedures , mouse handling , and husbandry were performed according to guidelines issued by the Institutional Animal Care and Use Committees ( IACUC ) at the University of Southern California and after approval from the institutional IACUC committee . The transgenic construct utilized by Park et al . [28] to test Six2-DE enhancer activity was modified to insert PacI and SwaI sites for cloning downstream of the Hsp68 minimal promoter followed by an IRES-NLS-GFP-BirA cassette . However , the IRES-NLS-GFP-BirA cassette was not included in the generation of the Six2-BF line . Each transcription factor of interest ( Six2 , Hoxd11 , and Osr1 ) was amplified from E15 . 5 kidney cDNA with a BioTag-3XFLAG sequence on the 3’ C-terminus and inserted into the transgenic vector using PacI and SwaI sites . Transgenes were purified and injected as previously described [28] . F0 animals were genotyped and transgenic animals bred to confirm germline transmission . Embryonic offspring were also analyzed for correct expression patterns of the transgene in the developing kidney . Cas9-mediated removal of the Six2-DE ( chr17:85747271–85749534 ) was performed by identifying optimal gRNAs flanking the enhancer utilizing the CRISPR Design Tool ( crispr . mit . edu; 5’ gRNA: gttaccatctacggtgatgc , chr17: 85747271–85747290; 3’ gRNA: gatatgattctcccgagctt , chr17: 85749515–85749537 ) . The gRNAs were cloned into the pX330-U6-Chimeric_BB-CBh-hSpCas9 plasmid ( Addgene ) as described ( ‘One-page Protocol for Cloning Using CRISPR Cas9 Backbone Plasmids’ found at http://www . genome-engineering . org/crispr/; [96] ) . Pronuclear injection of 2 . 5ng/μl of each construct into B6SJLF1/J x B6SJLF1/J zygotes ( The Jackson Laboratory ) with transfer to Swiss Webster ( The Jackson Laboratory ) pseudopregnant females was performed in house . A similar strategy was used to create the Wnt4-∆DE mouse ( chr4:137216986–137217756 ) . The CRISPR Design Tool ( crispr . mit . edu ) was utilized to identify optimal gRNAs flanking the enhancer ( 5’ gRNA: aggctgacaagcgaagttac , chr4:137216986–137217008; 3’ gRNA: atgtcggttgattaataatc , chr4: 137217756–137217778 ) . The following primers were used to generate complexes for in vitro transcription of the gRNAs using the MEGAshortscript T7 Transcription Kit ( Ambion ) : Six2-DE 5’ gRNA F: AATAATACGACTCACTATAAGGCTGACAAGCGAAGTTACGTTTTAGAGCTAGAAATAGC , Six2-DE 3’ gRNA F: AATAATACGACTCACTATAATGTCGGTTGATTAATAATCGTTTTAGAGCTAGAAATAGC , Wnt4-DE 5’ gRNA F: AATAATACGACTCACTATAGAGGCTGACAAGCGAAGTTACGTTTTAGAGCTAGAAATAGC , Wnt4-DE 3’ gRNA F: AATAATACGACTCACTATAGATGTCGGTTGATTAATAATCGTTTTAGAGCTAGAAATAGC , Common R: AAAAGCACCGACTCGGTGCCACTTTTTCAAGTTGATAACGGACTAGCCTTATTTTAACTTGCTATTTCTAGCTCTAAAAC . Cytoplasmic injection of 100ng of each gRNA and 50ng of Cas9 mRNA ( TriLink Biotech ) into B6SJLF1/J x B6SJLF1/J zygotes with transfer to Swiss Webster pseudopregnant females was performed in house . To establish lines , F0 animals were born and genotyped to confirm presence of the deletion . One animal carrying a deletion from chr17:85747284–85749542 for the Six2-DE and a deletion of chr4:137216991–137217771 for the Wnt4-DE ( each confirmed by Sanger sequencing of PCR product ) was mated to C57BL/6J ( The Jackson Laboratories ) to establish the line . Genotyping primers: Six2-DE flank F: GCAGAATGAGATTCTGACAGCCCAG Six2-DE flank R: CAAGGATGTCTTGTTTGGTCCTTGAGTGAG Six2-DE internal F: GAGGCCCATAAATAAAGCTGGGACG Six2-DE internal R: CTCCAGTGACAGATACCACTCTTACTG Wnt4-DE flank F: AAGCCATGAGGAAAAGAGGGTT Wnt4-DE flank R: TTCTCAACCCCAAACCCCACC Wnt4-DE internal F: AGTGTGAGGCACTGTGTAGC Wnt4-DE internal R: GTGGATGCTGCCTTATGGGT Six2TGCtg or Cited1-nuc-TagRFP-Ttg lines utilized for FACS were previously described [69] ( http://www . gudmap . org/index . html ) . Six2GCE , Six2CE , and Wnt4GCE mice were previously described [69] . The Br mutant mouse and mapping are previously described [97] and experimental protocols were approved by the University of Hawaii Institutional Animal Care and Use Committee . Wildtype or transgenic kidneys were utilized for ChIP . Hoxd11-BFtg/+ and Osr1-BFtg/+ kidneys were sorted out prior to ChIP by visualizing GFP+ kidneys . ChIP from E16 . 5 whole kidneys was carried out as previously described [3] . Nephron progenitor purification by MACS prior to ChIP was carried out is described in Brown et al . , 2015 [45] . Briefly , E16 . 5 mice kidney cortex cells were dissociated with collagenase/pancreatin as described in Brown et al . , 2015 [45] , fixed with 1% formaldehyde for 10 min in room temperature in AutoMACS running buffer . The fixed cells were then processed following published protocols [74] to generate 4C libraries . Dpn II and NlaIII were used in the first and second restriction enzyme digestions , respectively . In order to create a view point from Six2-DE , the primers used in 4C-PCR: Six2DE DpnII reading F: tccctacacgacgctcttccgatctGTTCTGAAAGAGCCGTGTAGGGATC Six2DE NlaIII noReading R: gtgactggagttcagacgtgtgctcttccgatcGGGGCCCATAAATCGTGATTCAAC The capitalized letters indicate the complimentary sequences to the genomic view point and the remainder Illumina adaptor sequences . The 4C-libraries were then indexed by PCR and sequenced by NextSeq500 . 4C-seq data were analyzed following the workflow provided by 4C-ker [98] . Briefly , the data was mapped to a reduced genome containing 25 bp regions from DpnII sites genome-wide and were subsequently quantified in 3 kb windows to show enrichment . The 4C-seq data is deposited on GEO ( GSE90017 ) . All ChIP-seq sequences were mapped to the mouse reference genome ( mm10 ) using Novoalign software ( Novocraft; parameters: single-end reads trimming 10 bp , polyclonal read filter: 7 , 10 0 . 4 , 2 , maximum alignment score acceptable: 120 ) . Mapped ChIP-seq and input data were analyzed using QuEST 2 . 4 software [99] using a “transcription factor” setting . The false discovery rate ( FDR ) for detecting the bound regions was evaluated by allocating the same number of mapped reads from a separate mouse input library and performing QuEST analysis using the same parameters . We generated multiple replicates for each ChIP-seq experiment ( except Hoxd11 due to technical issues ) , and used the replicate containing the most peaks using the same peak calling parameters for downstream analyses ( see S1G and S3A Figs , G for peak overlap between replicates . The smaller replicates all had >50% overlap with the larger replicate ) . To account for the innate differences between transcription factors in binding to the genome , we used different parameters in calling peaks of different transcription factors . We used high ratio of peaks with motif and low variability of motif-peak distances as our standard in determining the validity of a data set . See S7 Table for ChIP-seq data information and parameters in peak calling . In this paper , overlapping sites are defined as those with ChIP-seq peak center distance <150 bp from each other unless otherwise specified . To evaluate the statistical significance of two sets of peaks overlapping each other ( Fig 2A ) , we performed binomial test with the null hypothesis that peaks fall randomly into open chromatin regions in nephron progenitors . To determine the open chromatin regions in nephron progenitor , we performed ATAC-seq [100] in MACS-purified nephron progenitors [45] . We called ATAC-seq peaks with QuEST using the ‘transcription factor’ setting following threshold of fold enrichment > 10 . Then we extended the ATAC-seq peak coordinates by 150 bp to both sides , resulting in a total size of accessible chromatin as 21856500 bp . This process is modeled as following: NA , B∼Binom ( NA , pB ) pB=300∙NBSizeaccessiblegenome where NA , B is the number of peaks in set A that overlap with set B , and pB is the probability of a randomly located peak overlapping with set B . Information on all ChIP-seq samples presented in the paper can be found in S7 Table . The ChIP-seq and ATAC-seq data is accessible from GEO ( GSE90017 ) . All de novo motif discovery work was carried out using MEME ( Multiple Em for Motif Elicitation , [101] ) . To find the most enriched motifs for each transcription factor , MEME was run on a pool of 100 bp sequences around the predicted peak center for the top 1000 ChIP-seq peaks called from each data set ( or all peaks if number of the total peaks is less than 1000 ) . The locations of motif within 300 bp of peaks are found by FIMO ( Find Individual Motif Occurrences , [102] ) with data set-specific p-value threshold setting ( Six2 motif: 2e-4; Hoxd11 motif: 2e-4; Osr1 motif: 2e-4; Wt1 motif: 2e-5; bHLH motif: 2e-5 ) . We model the appearance of a motif near a set of peaks as following: NA , m∼Binom ( NA , pm ) pm=Nm300∙NA where NA , m is the number of motif m found in +/-150 bp of the peak set A; pm is the probability of finding such motif near a matched random set with the same number of peaks in A . Since promoter regions of genes are GC-rich , resulting higher rate of discovering GC-rich motif ( Wt1 , Osr1 and bHLH ) than TA-rich motif ( Six2 , Hoxd11 ) in promoter regions . To address this bias , we created the matched random set of peaks by picking genomic coordinates with the same distances to the nearest TSS as the observed peaks set , but with permutated nearest genes . We then screen the matched random peaks for the same motif to obtain Nm . GREAT GO analysis was performed utilizing the online GREAT program , version 2 . 0 [43] . Gene regulatory domains utilized for region annotation were defined as minimum 5 . 0 kb upstream and 1 . 0 kb downstream of the TSS , and extended up to 500 . 0 kb to the nearest gene’s minimal regulatory domain ( ‘single nearest gene’ option ) . GO Biological Processes annotations were assessed for each peak category . To infer the statistical significance of a set of ChIP-seq peaks found near a set of genes ( Fig 3 ) , we performed the following analysis . We assigned +/-500 kb from TSS of a gene as its ‘regulatory domain’ . We then calculated the probability of a selected set of ChIP-seq peaks falling into the merged regulatory domains of a list of selected genes . This is modeled by a binomial process with the null hypothesis that each peak falls uniformly throughout the genome . NA , x is the number of peaks in set A that fall into the regulatory domains defined by the gene list Gx . px is the probability of a peak falling into regulatory domains defined by the gene list Gx , assuming the peak randomly falling on any position in the genome . Size ( RRi ) is size of a regulatory region after resolving the overlap with any nearby regulatory regions . We found that each observed set of peaks fall into any random sets of regulatory domains more often than expected , which is not observed when doing the same experiment using random sets of genomic coordinates . To control this background over-representation , we obtained a background enrichment ratio over random sets of regulatory domains by rA , x=1n∑i=1nNA , xri/ ( NApxri ) where rA , x is the normalizing factor for a specific set of peaks A . NA , ri is the number of peaks from set A that fall into a the regulatory domain defined by a random list of genes Gxri which contains the same number of genes as Gx . Therefore , the final binomial model we used in the analysis is NA , x∼Binom ( NA , pxrA , x ) Cortical tissues of E16 . 5 or P2 kidneys from Six2TGCtg/+ or E16 . 5 Cited1-nuc-TagRFP-Ttg/+ embryos were dissociated as described in Brown et al . , 2015 [45] . The dissociated cells were resuspended in autoMACS buffer ( Miltenyi Biotec ) and passed through a 40 μm nylon filter to obtain single cells . The respective GFP+ , GFP- , or RFP+ cells were then isolated with the BD FACSAria II . RNA was isolated from FACS isolated cells using the QIAGEN RNeasy Micro Kit . RNA was submitted to the USC Epigenome Center for library preparation and sequencing on the Illumina HiSeq 2000 . All RNA-seq reads were aligned to the mouse reference genome ( mm10 ) using the TopHat2 [103] . Sequences have been deposited in GEO , accession number GSE90017 . Quantification of RNA-seq reads to generate RPKM was performed by Partek Genomics Suite software , version 6 . 6 ( St . Louis , MO , USA ) . TPM was calculated by dividing the RPKM by the mapping ratio of the library to exon regions of the genome . To identify genes differentially expressed in a cell type , we select those with a fold difference > 3 , TPM > 5 and p-value < 0 . 05 . Sample information can be found in S7 Table . A complete list of all annotated genes and their coordinating RNA-seq data can be found in S3 and S4 Tables , and coordinating ChIP-seq data can be found in S8 Table . Gene ontology analysis of gene lists was carried out by PANTHER [104] . For the gene set analysis in Fig 3D , we selected the enriched gene lists using the following metrics: nephron progenitor-enriched ( TPM > 10 in E16 . 5 Six2GFP+ cells and fold change > 2 in E16 . 5 Six2GFP+ vs . Six2GFP- cells ) , self-renewing nephron progenitor-enriched ( TPM > 10 in E16 . 5 Cited1RFP+ cells and fold change > 2 in E16 . 5 Cited1RFP+ vs . P2 Six2GFP+ cells ) and differentiating nephron progenitor-enriched ( TPM > 10 in P2 Six2GFP+ cells and fold change > 2 in P2 Six2GFP+ vs . E16 . 5 Cited1RFP+ cells ) . S5 and S6 Tables are pre-filtered to show the nephron progenitor-enriched genes only , but entire lists can be viewed by releasing the filter . qPCR reaction was performed with Luna Universal qPCR Master Mix Protocol ( New England Biolabs ) on a Roche LightCycler 96 System . The primers used in this paper includes: GAPDH F: AGGTCGGTGTGAACGGATTTG GAPDH R: TGTAGACCATGTAGTTGAGGTCA Six2 F: CACCTCCACAAGAATGAAAGCG Six2 R: CTCCGCCTCGATGTAGTGC Pax2 F: AAGCCCGGAGTGATTGGTG Pax2 R: CAGGCGAACATAGTCGGGTT Wnt4 F: AGACGTGCGAGAAACTCAAAG Wnt4 R: GGAACTGGTATTGGCACTCCT In situ hybridizations were performed on frozen sections as previously described ( https://www . gudmap . org/Research/Protocols/McMahon . html ) . The primer sequences used to generate Wnt4 probe template are: F: GAGAAACTCAAAGGCCTGATCCA R: TAATACGACTCACTATAGGGGGCTTTAGATGTCTTGTTGCACG EMSA was carried out using Glutathione S-transferase ( GST ) -tagged recombinant proteins purified from bacterial lysates . To produce the protein , bacterial expression constructs ( pDEST15 backbone ) were prepared using the Gateway system . BL21-AI One Shot ( Life Technology ) chemically competent cells were transformed and grown to OD600 = 0 . 6 before induction with 0 . 2% L- ( + ) -arabinose ( Sigma ) for 3 hrs at 37 °C . The bacteria pellets were resuspended in lysis buffer ( 20 mM Tris-HCl , 150 mM NaCl , 1% TritonX-100 , 1x protease inhibitor , 5 mM DTT ) and incubated with 1 mg/mL lysozyme ( Sigma ) . The lysates were sonicated with Branson digital sonifier at 50% amplitude for 90 s . After sonication , supernatant of the lysates were incubated with 5mL glutathione-agarose beads ( Sigma ) per 1 L bacteria culture for 1 hr at 4°C . The beads were washed with 1% TritonX-100/PBS and eluted with elution buffer ( 20 mM Tris-HCl , 150 mM NaCl , 15 mg/mL ( 50 mM ) reduced glutathione , 1x protease inhibitor ) . The eluted protein was concentrated to at least 10 mg/mL using Amicon Centrifugal Filter Unit with the right filter size . The concentrated protein was diluted with PBS and concentrated again to exchange buffer . To perform the EMSA experiments , the recombinant protein was incubated with biotinylated DNA probe for 30 min at room temperature , then the mixture was run through a native TBE gel . The gel was transferred to a nitrocellulose membrane , which was then illuminated using the LightShift Chemiluminescent EMSA Kit ( Pierce ) . The sequences of DNA probes can be found in S2 Fig . Nuclear lysates were prepared from E16 . 5 whole kidneys using the Nuclear Complex Co-IP Kit ( Active Motif ) . Normal rabbit IgG or Six2 ( Proteintech , 11562-1-AP ) antibodies were crosslinked with dimethyl pimelimidate to Dynabeads Protein G ( Thermo Fisher Scientific ) using the Protein A/G SpinTrap Buffer Kit ( GE Healthcare ) . Nuclear extracts were incubated overnight with beads at 4°C . Samples were washed 5x with TBS+0 . 1% Triton X-100 , and proteins subsequently eluted with 0 . 1M Glycine-HCl , pH 2 . 9 . Samples were run on a 10% SDS-PAGE gel , transferred to nitrocellulose , and subjected to standard Western blotting protocols using Six2 ( Proteintech ) , Hoxd11 ( Abcam , ab60715 ) , or Wt1 ( Santa Cruz , sc-192 ) antibodies . Kidneys were isolated at the appropriate stage and fixed in 4% PFA for 1 hour . Cryosections were immunostained as previously described [28] . Antibodies used include Six2 ( Proteintech , 11562-1-AP ) , FLAG ( Sigma , F1804 ) , Wt1 ( Abcam , ab89901 ) , pan-cytokeratin ( Sigma , C2931 ) , Pax8 ( Abcam , ab13611 ) , Ecad ( Sigma , U3254 ) , Six3 ( Rockland , 200-201-A26S ) , and LTL-FITC ( Vector Labs , FL-1321 ) . Images were acquired on a Nikon Eclipse 90i epi-fluorescent microscope or Zeiss LSM 780 inverted confocal microscope . Purified genomic DNA from one wildtype and two Br/Br animals was sequenced on the Illumina Hi-seq 2000 and mapped to the mm10 genome using Bowtie2 . Specific details of the mapping and allele characterization are described further in S1 Supporting Information . Sequences have been deposited in GEO , accession number GSE90017 . | Nephrons , the filtering units of the kidney , derive from nephron progenitors . Deficiencies in nephron number increases the risk of kidney disease . An understanding of the regulatory programs governing progenitor actions has important translational potential . Several transcription factors regulate the nephron progenitor population . However , their target interactions are largely unknown . Here , we mapped and intersected the genome-wide binding sites for four such factors in mouse nephron progenitor cells in the developing kidney: Six2 , Hoxd11 , Osr1 , and Wt1 . The intersectional data highlight a high-value set of putative enhancer elements linked to genes regulating nephron progenitor properties . We validate the function of two such enhancer elements regulating the levels of Six2 , a key transcriptional regulatory factor in nephron progenitor maintenance , and Wnt4 , a critical signaling factor controlling the mesenchyme to epithelial transition of induced nephron progenitors . Further characterization of the Six2 regulatory landscape identified higher order regulatory interactions that ensure appropriate enhancer-promoter specificity . CTCF-bound sites between Six2 and the adjacent Six3 locus likely act as boundary elements to define topological interactions domains separating enhancer elements thereby providing distinct tissue specificity to each gene’s expression . An inversion of this region in the Brachyrrhine ( Br ) mutant mouse reverses Six2 and Six3 expression domains , placing Six3 under control of the Six2 enhancer element above resulting in kidney-specific expression , while Six2 expression shifts to the lens , a normal expression domain for Six3 . Together , these data expand our view of the regulatory genome and regulatory landscape underpinning mammalian nephrogenesis . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"medicine",
"and",
"health",
"sciences",
"gene",
"regulation",
"regulatory",
"proteins",
"dna-binding",
"proteins",
"nephrons",
"transcription",
"factors",
"sequence",
"motif",
"analysis",
"mammalian",
"genomics",
"kidneys",
"research",
"and",
"analysis",
"methods",
"sequence",
"analysis",
"transcriptional",
"control",
"bioinformatics",
"proteins",
"gene",
"expression",
"animal",
"genomics",
"biochemistry",
"anatomy",
"database",
"and",
"informatics",
"methods",
"genetics",
"biology",
"and",
"life",
"sciences",
"renal",
"system",
"genomics"
] | 2018 | Transcriptional regulatory control of mammalian nephron progenitors revealed by multi-factor cistromic analysis and genetic studies |
Although the pharmaceutical industry's “neglect” of neglected tropical diseases ( NTDs ) has been investigated , no study evaluating media coverage of NTDs has been published . Poor media coverage exacerbates the neglect . This study aimed to investigate , describe , and analyse international media coverage of “neglected diseases” in general and three specific NTDs—African trypanosomiasis , leishmaniasis , and Chagas disease—from 1 January 2003 to 1 June 2007 . Archives of 11 leading international , English-language media were searched . A content analysis was done , coding for media organisation , date , author , type of report , slant , themes , and “frames” . Semi-structured interviews with journalists and key informants were conducted for further insight . Only 113 articles in a 53-month time period met the inclusion criteria , with no strong trends or increases in coverage . Overall , the BBC had the highest coverage with 20 results , followed by the Financial Times and Agence France Presse . CNN had the least coverage with one result . The term “neglected diseases” had good media currency and “sleeping sickness” was far more widely used than trypanosomiasis . The disease most covered was leishmaniasis and the least covered was Chagas . Academic researchers were most commonly quoted as a main source , while the World Health Organization ( WHO ) and pharmaceutical industry were the least quoted . Journalists generally agreed NTDs had not been adequately covered , but said a lack of real news development and the need to cater to domestic audiences were major obstacles for NTD reporting . All journalists said health agencies , particularly WHO , were not communicating adequately about the burden of NTDs . Public health agencies need to raise priority for NTD advocacy . Innovative strategies , such as reporting grants or creating a network of voices , may be needed .
Roughly one in six people globally , mostly the very poor , suffer from one or more NTDs . [1] These diseases may not directly result in high mortality rates , yet cause much morbidity , suffering and poverty . [2] Despite this , NTDs are a low priority for the pharmaceutical industry , lacking safe and effective treatments; are overlooked by mainstream global health efforts , receiving little funding; and are ignored by the media , rarely making headlines . Even public health authorities have downplayed NTDs – often , they are not perceived as health burdens and do not require compulsory reporting . [2] In recent years , there has been a surge of activity around NTDs . The Drugs for Neglected Diseases Initiative ( DNDi ) , kickstarted by Médecins Sans Frontières ( MSF ) , and the Institute of OneWorld Health ( IOWH ) , were both set up to help spur development of drugs . Through public-private partnerships , new drug projects have flourished , with 63 ongoing by the end of 2004 . [3] The drug gap from market failure has been studied . A 2006 study found that in the past 30 years , only10 drugs were marketed for “most neglected diseases”; ( this figure rises to 21 if malaria and tuberculosis drugs are included ) . [4] However , NTDs and the media have not been studied . News reportage has been described as a “significant background” to policy change . [5] The importance of media advocacy in pushing forward tobacco control objectives has been demonstrated in studies . [6] It is thus timely and appropriate that greater attention be given to NTD advocacy . This study aims to investigate , describe and analyse international media coverage of “neglected diseases” in general and three specific NTDs – African trypanosomiasis , leishmaniasis and Chagas disease ( also known as American trypanosomiasis ) – between 01 January 2003 and 01 June 2007 . These parasitic diseases were chosen as they are some of the most neglected diseases , affecting people in three continents . The study period was timed around a key DNDi NTD advocacy campaign to ascertain whether the campaign had influenced media coverage . The study aims to provide a context of the current media situation facing NTDs and help future advocacy work .
The media selected were BBC online , CNN . com , the international news wire Agence France Presse ( AFP ) , the American news magazine Time , the international news magazine The Economist , the international business paper Financial Times , two British newspapers – The Guardian and Daily Telegraph – and three American newspapers – The New York Times , Washington Post and Los Angeles Times . Their databases are available online , some with a fee for access . This selection was made as they have: Articles were defined as focussing on NTDs in general or one of the three NTDs studied if they had: The search terms included the term “neglected diseases” , medical names of the three diseases plus the names “sleeping sickness” , “kala azar” and “black fever” ( a literal translation of “kala azar” ) , which has been used in the US media without mention of any other disease name . Articles with only one mention of the term “neglected diseases” ( in one paragraph ) were excluded from the analysis but recorded separately to note how many times this term was used . The coding system used was adapted from methodological frameworks used in other content analyses , particularly to track tobacco coverage . [8] [9] Articles were categorized by disease and media organisation , to note what diseases were covered and where , with a “general” category for articles discussing more than one ND but none in particular . Articles were also identified by author ( if available ) , date , type of report ( such as editorial or feature ) and slant of reporting ( negative , neutral or positive to NTD advocacy objectives ) . Qualitative analysis involved identifying topics and the “framing” of issues . Frame analysis has been described as a “means of explaining the ways that dominant news discourses evolve and come to define… a problem” . [5] Semi-structured interviews were performed with nine leading health journalists to gain insight into the findings and investigate factors influencing reporting . Journalists were chosen from leading media organisations such as BBC , CNN , Reuters , AFP and Associated Press . Leading global health journalists from the Financial Times ( FT ) , the Boston Globe and Washington Post were also interviewed . One academic and three former journalists now working on advocacy for international health agencies ( two formerly with WHO ) were also interviewed .
The BBC had the highest coverage with 20 results , followed by the FT and AFP with 19 and 18 results respectively ( see Table 1 ) . CNN had the least coverage with one result , for a story originally from Reuters . There were a wide variety of articles in the BBC and a notable number on sleeping sickness , including , uniquely , some from the field featuring patients . The FT had the most in-depth and detailed articles , often with exclusive financial perspectives . The new business model offered by public-private partnerships was explored in detail , in view of their potential for underfunded areas . The FT actually had the largest number of articles but many discussed the issue generally , rather in specific diseases , so did not meet the inclusion criteria . Although AFP's articles covered a wide area , reflecting the agency's mission , with three articles on sleeping sickness in Africa , none focussed on Chagas disease . Many stories from other organisations had a strong domestic angle , such as interviews with British scientists working on NTDs . The work of the US-based organisation IOWH , “black fever” among US troops in Iraq and the presence of Chagas in the US blood supply were some topics in the New York Times . The Economist offered long , in-depth analyses while none of the articles in Time or Daily Telegraph were “hard news” stories . The term “neglected diseases” was commonly used , which partly explained why many articles fell in the “general” category rather than a specific disease . The disease most covered was leishmaniasis , mainly because of the wide reach of those affected ( which includes US troops in Iraq ) and recent drug developments led by IOWH . African trypanosomiasis was the next most covered disease ( as sleeping sickness ) , primarily by the BBC . Chagas had significantly less coverage with no articles in the British media other than one in the BBC . The main focus of Chagas by the American media was the parasite's threat to the American blood supply . No article actually focussed on the problem itself in South America . The most common group to be quoted were local university researchers , accounting for main quotes in a third of articles with quotes . Academics represent a local , accessible and relatively independent source . With the “medical researchers” group , they accounted for 41% of all main quotes . WHO was quoted as a main source in only 4% of articles , while the Bill and Melinda Gates Foundation and pharmaceutical industry had even poorer media visibility . DNDi and MSF accounted for 18% of all main quotes . It was difficult to identify clear “frames” , but some broad themes of focus did emerge . The general need for more attention on NTDs , including calls for more research , drugs and funding , was the most common theme of articles collectively , with 19 articles . Public-private partnerships – which included the work of IOWH , DNDi and other such institutions - were another common focus , with 15 articles . Frames depicting the horror or tragedy of NTDs , often describing the reality of these “forgotten” diseases in term of the epidemic , victims , drugs or situation were almost as common . Other focuses for articles included: Journalists generally agreed that NTD were an important story that had not been adequately covered , but with the caveat that news stories had to be “newsworthy” . Health coverage veered towards “breaking news” such as bird flu outbreaks –headline-hitting events that raised ratings . Journalists who did cover NTDs were often personally motivated . Andrew Jack of the FT , who had the largest number of bylines in the study , said his reporting was “100%” driven by his interest . A lack of real news development , the drive to cater to domestic audiences and competing health interests were cited as the main obstacles for NTD reporting . “Poor people dying from an illness is not news , ” unless there is some change or development , one producer from an international broadcaster said . But HIV/AIDS was widely reported on “because it sells stories” and has the funding and attention of policymakers . Coverage of global health issues was particularly poor in the American media , where health and foreign budgets are facing cuts . All journalists said health agencies were not communicating adequately about the burden of NTDs . Some journalists were particularly critical of WHO and the Bill and Melinda Gates Foundation for the difficulty in reaching officials for comment . Bill Gates was , however , credited with raising the profile of NTDs . NGOs such as MSF were cited as good sources for stories . Journalists said stories needed to have a broad appeal which touched core readership to get covered . New developments or “breakthroughs” were easier to sell as stories . The “human element” was powerful , but few journalists were able to get such stories first-hand from the field . This represented a real constraint for coverage . One communications advisor ( consulting for DNDi ) said health agencies needed to present stories featuring “real people” rather than “experts in their ivory towers” and the “yuck” factor about these diseases needed to be played up to “grab the public imagination” rather than facts about the lifecycle of the parasite .
This study shows the general lack of coverage of NTDs in the media , with an average of about 10 articles per media organisation in a period of more than four years . By comparison , an unfiltered search for HIV/AIDS on AFP's database found more than 1 , 000 articles for the same period . There was a wide disparity in coverage between various media , with results for BBC 20-fold higher than CNN . No events or developments seemed to capture media interest across the board . The “newsworthy” element of NTDs clearly varied between media , ranging from the financial angles used by the FT to the emotive human stories featured in BBC . The results reflect select international media – other leading organisations ( such as Reuters news agency and the Wall Street Journal ) had to be excluded due to time constraints . Also , only English-language media were selected and the term “international” is somewhat debatable . It would be useful to repeat the study in other language and compare ND coverage; a brief survey of Le Monde found many NTD articles , particularly on DNDi . Also , although some media had more profound articles on NTDs , this was not analysed due to a lack of an objective coding variable . However , this investigation is the first study to systematically analyse NTD media coverage . Further , the media selected still represent a sample of key media and some patterns clearly emerged . For example , the penchant for a local angle was even parochial at times . Stories get written about leishmaniasis in pets before humans , as was seen in The Daily Telegraph . In the time frame of the study , activity by celebrities and the Global Network for NTDs ( gnntdc . org ) , did not yet result in coverage by the mainstream media included in the study . However this may improve as celebrity activity and the networks pick up more media currency , especially with the impetus provided by President George Bush's 2008 NTD Initiative . The interviews provided much insight , particularly on the struggle to cover global health issues in the American media , where foreign news budgets have been slashed . [10] One study found US foreign news coverage on front pages fell significantly from 1987 to 2004 , from 27% to14% . [11] Interestingly , the news organisations with the first and third most coverage ( BBC and AFP ) both receive some public funding , so do not operate on an entirely commercial basis . It is under such a challenging context that journalists face the pressure of reporting on relatively unknown diseases with limited information . Added to this is the difficulty in getting information on NTDs . Providing ready access to information and experts when needed is critical to improve coverage . Forming coalitions or networks could also help strengthen voices in the media . In selling a story , terminology was enormously important . “Human African trypanosomiasis” was clearly off-putting for journalists , who overwhelmingly preferred the term “sleeping sickness” . Journalists also found “neglected diseases” more catchy and concise than “neglected tropical diseases” . Clearly , NTD advocates need to speak the same language as journalists to engage the media . The study also showed the lack of vivid and powerful “human” stories from the field ( very few stories quoted patients ) which generally have media appeal . One solution would be for NGOs to sponsor journalists to join them in the field , but this may raise the thorny issue of independent reporting . In the market-driven setting of today's media , more innovative strategies may be needed . The same commercial context that constrains drug development of NTDs also curbs global health reporting , particularly in the American media . Just as public-private partnerships have transformed the landscape of drug development , some public-private funding may be needed to bring insightful , in-depth reporting on NTDs from the field to the pages of Western newspapers . Many fellowships , grants and awards are already available to promote reporting in certain fields . The Kaiser Family Foundation supports HIV/AIDS reporting projects [12] and offers international health fellowships while Harvard University recently started Nieman fellowships in global health reporting , with a US$1 million grant from the Bill and Melinda Gates Foundation . [13] This study showed that even in a select group of media , there are clear patterns in what diseases get covered , what topics , terms and sources are preferred , and in which media . The disparity in coverage between media reflects different news priorities and interests , yet also opens a door to potentially increasing coverage , particularly amidst growing interest in global health . Public health agencies need to consider sustained and innovative advocacy on NTDs . A variety of strategies may be needed , including those to shift current “frames” – media portrayal and perception of NTDs . | In recent years , there has been a flurry of activity to reverse the neglect that has characterised NTDs , mostly focussed on drug development . The drug gap may be explained by market failure , yet other forces also conspire to cause the neglect of NTDs . One problem is the low visibility of these diseases . By comparison , the high-profile “big three” infectious diseases of AIDS , tuberculosis , and malaria have received increased donor attention and funding with greater visibility . Efforts to remove the “neglect” from NTDs must involve raising their profile . This study , focussing on three of the most neglected diseases , aims to provide a context of the current media situation—the what , where , and why of NTD coverage—to support future advocacy work . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"infectious",
"diseases/neglected",
"tropical",
"diseases",
"public",
"health",
"and",
"epidemiology"
] | 2008 | Neglected Diseases in the News: A Content Analysis of Recent International Media Coverage Focussing on Leishmaniasis and Trypanosomiasis |
Bats are newly identified reservoirs of hantaviruses ( HVs ) among which very divergent HVs have been discovered in recent years . However , their significance for public health remains unclear since their seroprevalence as well as antigenic relationship with human-infecting HVs have not been investigated . In the present study archived tissues of 1 , 419 bats of 22 species from 6 families collected in 5 south and southwest provinces in China were screened by pan-HV RT-PCR following viral metagenomic analysis . As a result nine HVs have been identified in two bat species in two provinces and phylogenetically classified into two species , Laibin virus ( LAIV , ICTV approved species , 1 strain ) and Xuan son virus ( XSV , proposed species , 8 strains ) . Additionally , 709 serum samples of these bats were also analyzed by ELISA to investigate the seroprevalence and cross-reactivity between different HVs using expressed recombinant nucleocapsid proteins ( rNPs ) of LAIV , XSV and Seoul virus ( SEOV ) . The cross-reactivity of some bat sera were further confirmed by western blot ( WB ) using three rNPs followed by fluorescent antibody virus neutralization test ( FAVNT ) against live SEOV . Results showed that the total HV seropositive rate of bat sera was 18 . 5% ( 131/709 ) with many cross reacting with two or all three rNPs and several able to neutralize SEOV . WB analysis using the three rNPs and their specific hyperimmune sera demonstrated cross-reactivity between XSV/SEOV and LAIV/XSV , but not LAIV/SEOV , indicating that XSV is antigenically closer to human-infecting HVs . In addition a study of the distribution of the viruses identified an area covering the region between Chinese Guangxi and North Vietnam , in which XSV and LAIV circulate within different bat colonies with a high seroprevalence . A circulation sphere of bat-borne HVs has therefore been proposed .
Hantaviruses ( HVs ) , members of the genus Orthohantavirus within the family Hantaviridae in the order Bunyavirales , are responsible for two major life-threatening diseases in humans: hemorrhagic fever with renal syndrome ( HFRS ) in Eurasia and hantavirus cardiopulmonary syndrome ( HCPS ) in the Americas [1] . Every year around 100 , 000 HFRS cases and 1 , 000 HCPS cases are reported worldwide [2] . China suffers severely from epidemic HFRS; in 2017 alone , official statistics reported 11 , 262 cases with 64 deaths [3] . HVs are predominantly carried and transmitted by rodents , but insectivores and bats have also been reported as hosts . Several bat-borne HVs are presently known , which show large genetic diversities from currently known rodent- and insectivore-borne HVs . The first reported bat-borne HVs , Magboi virus ( MGBV ) and Mouyassué virus ( MOUV ) , were identified respectively in Sierra Leone and Côte d’Ivoire of Africa in 2012 [4 , 5] . Then two bat-borne HVs , Longquan virus ( LQUV ) and Huangpi virus ( HUPV ) , were reported in China in 2013 [6] , followed by the detection of Xuan son virus ( XSV ) at three locations in North Vietnam [7 , 8] . We reported the first complete genome of a bat-borne HV , Laibin virus ( LAIV ) , identified from a black-bearded tomb bat in Guangxi Province of China in 2015 [9 , 10] . Since then three more complete genomes of bat-borne HVs , Makokou virus ( MAKV ) , Quezon virus ( QZNV ) and Brno virus ( BRNV ) have been reported sequentially in 2016 , in Central Africa ( Gabon ) , Southeast Asia ( Philippines ) and Central Europe ( Czech Republic ) , respectively [11–13] . Most recently , a sister lineage of MOUV was detected in dried blood samples from bats in Eastern Africa ( Ethiopia ) in 2017 [14] . Of these HVs only three , Laibin , Longquan and Quezon viruses were approved as bat-borne HV species within genus Orthohantavirus in the 10th report of International Committee on Taxonomy of Viruses ( ICTV ) released in 2017 [15] . Phylogenetic analysis of bat-borne HVs has indicated that bats might be the natural original hosts of HV: i . e . , the viruses first appeared in bats or insectivores , then emerged in rodents [6 , 16–19] . However , due to lack of sufficient bat-borne HV genomic sequences , their evolutionary phylogeny and genetic diversity as well as biological features are poorly understood . HVs are enveloped and spherical in shape although pleomorphic forms are also found with the diameters ranging from 80–120 nm . Within the capsid is a tripartite negative-stranded RNA genome consisting of small ( S ) , medium ( M ) and large ( L ) segments with a total length of about 11 . 8 kb , respectively encoding nucleocapsid protein ( NP ) , glycoprotein ( GP , a precursor for two viral surface glycoproteins , Gn and Gc ) and RNA-dependent RNA polymerase ( RdRp ) [1] . The NP is multifunctional and plays an essential role in viral replication , not only binding viral RNA strands to form a ribonucleoprotein ( RNP ) to prevent RNA from degradation , but also regulating virus replication and assembly [1 , 20 , 21] . NP is also the main target for the earliest immune response . Its coding gene is much more conserved than the GP gene , and is therefore commonly used as a diagnostic antigen for HV detection [22–25] . Different serotypes of HVs can be determined by an at least four-fold difference in two-way cross neutralization tests , and it has been reported that serotype-specific as well as group-common and genus-common epitopes can be found in the NP . Cross-reactivity has been found between different serotypes of HVs in rodents and insectivores [26–28] . However , the serological and antigenic relationships between bat- and rodent- or insectivore-borne HVs have not yet been studied . South and southwest China have a high density of bat population consisting of a large number of diverse species . Recently , investigations on bat viruses in this area have revealed many novel viruses , such as coronaviruses [29–34] , filoviruses [35 , 36] and group A rotaviruses ( RVA ) [37 , 38] . Among these , some bat-borne coronaviruses [29 , 30 , 33] and RVAs [38] have been found to cross species , causing outbreaks of emerging infectious diseases in human and pigs . South and southwest China are also the major epidemic areas of HFRS with all transmission events associated with exposure to rodents [39] . Although increasing number of HVs have been identified in bats , no investigation has been shown their seroprevalence and antigenic characters . The implication of bat-borne HVs to public health is still unclear . In present study , we have conducted systematic etiological and serological investigation on bat-borne HVs in south and southwest China , revealing the antigenic relationships between bat-borne and human-infecting HVs and identifying a geographic region between southwest China and north Vietnam in which divergent bat-borne HVs circulate .
Following viral metagenomic analyses of bat intestines and lungs , 18 contigs with lengths of 110–726 nucleotides ( nt ) from Laibin ( LB ) , Baise ( BS ) and Pu’er ( PE ) cities ( see S1 Fig ) were annotated to HVs . The highest nt identities of these ( 72–99% ) were shared with the M or L segments of LAIV or XSV ( S1 Table ) . By pan-HV PCR screening of all intestines and lung tissues , only lung tissue of 9 bats were found to be positive . One of 39 ( 2 . 6% ) black-bearded tomb bats ( Taphozous melanopogon ) in BS , Guangxi , 5 of 55 ( 9 . 1% ) pomona roundleaf bats ( Hipposideros pomona ) in LB , Guangxi , and 3 of 40 ( 7 . 5% ) pomona roundleaf bats in PE , Yunnan were also positive ( Table 1 ) . The viral sequence determined in black-bearded tomb bats in BS showed a 97% nt identity with previously reported LAIV BT20 [9] , and were therefore considered an LAIV variant , BT33 . The rest eight from LB and PE showed 93% and 82% nt identities with XSVs identified from Vietnam [7 , 8] , indicating they were all XSV variants and therefore respectively named XSV AR18 , AR19 , AR23 , AR28 , AR30 , PR10 , PR15 and PR30 . To isolate infectious viruses , homogenized lung tissues of five XSV ARs and three XSV PRs were separately pooled . The two pooled samples , along with one LAIV BT33 positive lung tissue were thoroughly homogenized by grinding and their filtered supernatants were incubated with African green monkey kidney ( Vero ) and the E6 clone and baby hamster kidney ( BHK-21 ) cell cultures . During five passages , no CPE was observed , and RT-PCR analyses of all passaged cultures were negative , with no HV isolated . To gain genetic insight into the HVs , the full genomes of LAIV BT33 , XSV AR18 , AR23 and PR15 were sequenced and analyzed using previously reported methods [9 , 40] . As shown in Table 2 , three gene segments of LAIV BT33 had the same sizes as previously reported LAIV BT20 [9] . Three segments of XSV AR18 and AR23 had exactly the same size ( 1 , 753 nt of S , 3 , 751 nt of M and 6 , 521 nt of L ) , while PR15 had similar sized S ( 1 , 743 nt ) and L ( 6 , 522 nt ) , but its M segment was shorter ( 3 , 584 nt ) than those of XSV AR18 and AR23 , resulting from a 50-aa deletion at the 5’ terminal of the coding sequence , corresponding to 6–55 aa of Gn protein . This deletion was confirmed by repeated RT-PCR and sequencing . Currently the function of Gn is largely unknown and 1–17 aa is the signal peptide of Gn responsible for translocation of Gn to Golgi [41 , 42] , therefore the deletion may have impact on Gn location . In addition , the highly conserved motif WAASA ( polyprotein-recognized pentapeptide ) in the M segment of HV was observed in all four strains , but the ORF in the S segment of some HVs ( such as Puumala , Tula and Andes viruses ) encoding a 7–12 KDa nonstructural protein ( NSs ) which functioned as an interferon antagonist were not found [43–45] . Sequence comparison of the four strains with other bat-borne HVs available in GenBank ( Table 2 ) showed that LAIV BT33 shared the highest ( 98 . 4–98 . 6% nt and 99 . 2–100 . 0% aa ) identities with LAIV BT20 in its three genomic ( full-length ) segments and low identities with other bat-borne HVs , ( 49 . 6–75 . 4% nt and 45 . 8–87 . 3% aa identities in full or partial gene segments ) , indicating that it is a variant of LAIV . XSV AR18 and AR23 shared the highest ( 91 . 8–93 . 4% nt and 99 . 0–100 . 0% aa ) identities with the XSV strain F42682 ( partial gene segments , full-length not available ) and XSV PR15 the highest ( 82 . 8–84 . 9% nt and 97 . 9–99 . 1% aa ) identities with XSV F44601 ( partial gene segments , full-length not available ) , indicating that they are novel variants of XSV . Full-length genomic sequence comparison of the four strains with those of rodent-and insectivore-borne HVs showed that bat-borne HVs of the present study had very low nt ( 43 . 3–66 . 6% ) and aa ( 40 . 0–67 . 6% ) similarities to rodent-and insectivore-borne HVs ( S3 Table ) . To construct phylogenetic relationships , 92 , 63 and 40 complete coding sequences of hantaviral NP , GP and RdRp respectively were used . Currently there are only 14 NP , 9 GP and 7 RdRp complete sequences of bat-borne HVs available in GenBank . As shown in Fig 1 , rodent- and insectivore-borne HVs , except for Nova virus ( NVAV ) [18 , 46] and Altai virus ( ALTV ) , showed a similar topology , which classified them within clades I , III and IV in all three trees . Bat-borne HVs showed different topology structures , however , which were all clustered together within clade II in the NP tree , but within three clades ( II , V and VI ) in the GP tree , or within two clades ( II and VI ) in the RdRp tree . It is interesting to note that two insectivore-borne HVs , NVAV and ALTV , respectively identified in Talpa moles and Sorex shrews , were genetically closer to bat-borne than to insectivore-borne HVs in the NP and GP trees [18 , 46] . To characterize the antigenic relationship between bat- and human-infecting HVs the entire NPs of LAIV BT33 , XSVAR18 and SEOV GuangzhouRn36 were expressed in E . coli and purified . Polyclonal anti-serum against the rNPs ( named anti-L , anti-X and anti-S , respectively , for LAIV , XSV and SEOV ) were prepared by immunization of mice , resulting in titers by ELISA of 8 , 000× , 4 , 000× and 4 , 000× respectively . Western blot ( WB ) analyses showed that SEOV-convalescent human serum ( H anti-SEOV ) had a significant cross-reactivity with the rNP of XSV but not LAIV , as well as the rNP of SEOV ( Fig 2A ) . To further characterize this cross antigenicity , three NPs were eukaryotically expressed with an EGFP tag . Further WB analyses with the three anti-rNP sera showed that anti-L reacted strongly with the eukaryotic rNP of XSV but not with that of SEOV , in addition to a very strong reactivity with its own LAIV rNP ( Fig 2B ) . In contrast , anti-S showed only a weak cross-reactivity with eukaryotic rNP of XSV and not at all with that of LAIV , although with very strong reactivity with its own SEOV NP . It is interesting to note that anti-X had cross-reactivity with eukaryotic rNPs of both LAIV and SEOV ( weak for LAIV and strong for SEOV ) , in addition to a very strong reactivity with its own XSV rNP . This result unexpectedly showed that bat-borne HVs do share cross reactivity with human-infecting HVs and that significant differences of antigenicity do exist in different bat-borne HVs . In our study XSV was antigenically closer to SEOV than LAIV . Results of the serological assay of 709 bat sera by ELISA against the three viruses are shown in S2 Fig with 88 of them being further confirmed by WB ( S3 Fig ) . Since no standard bat sera ( either positive or negative ) were available , the highest coincidence rate ( CR ) between WB and ELISA was used to determine OD492 ELISA positive cut-off values: 0 . 10 , 0 . 10 and 0 . 11 at the highest CR value for each virus ( 87 . 5% , 86 . 4% and 86 . 4% , respectively , for LAIV , XSV and SEOV ) ( see S4 Fig ) . With such cut-offs , the κ test showed high levels consistence between the two methods with Z values being 7 . 0862 for LAIV , 6 . 8255 for XSV , and 6 . 9270 for SEOV ( all p<0 . 0001 ) , and the high κ values being 0 . 7260–0 . 7505 . These results indicate that the established ELISA was valid to test the bat sera . Using these cut-offs , 131 of 709 ( 18 . 5% ) bat sera were found to be HV antibody positive . Fig 3 shows the distribution of OD492 readings of the 131 positive bat sera , with most sera having OD492 readings between the cut-off and 0 . 30 . To further determine antibody titers , the positive sera were 4-fold diluted from 100× to 1 , 600× and retested by ELISA . Results showed that most positive sera had titers of 100× , yet 18 sera reached 400× , with the H anti-SEOV at 1 , 600× ( Fig 4 ) . Of 131 positive sera , 55 ( 7 . 76% ) showed cross-reactivity to all three viruses , 19 ( 2 . 7% ) to both of LAIV and XSV , 9 ( 1 . 3% ) to both XSV and SEOV , and 7 ( 1 . 0% ) to both LAIV and SEOV , whereas sera reacted exclusively with one virus were only 9 ( 1 . 3% ) with LAIV , 10 ( 1 . 4% ) with XSV and 22 ( 3 . 1% ) with SEOV . This further showed that seroprevalence of HVs in bats widely existed in four provinces ( in Guangdong bat sera were not collected ) . As shown in Fig 5 , among 13 cities with bat serum collection 12 were seropositive with levels from 5 . 5% to 35 . 9% . Of 16 bat species tested 13 had seropositive rates ranging from 4 . 8% to 50 . 0% . Of 88 bat sera tested by WB , 48 with sufficient volume were further tested for neutralizing antibody ( NAb ) titers against SEOV by the fluorescent antibody virus neutralization test ( FAVNT ) . Results showed that nine bat sera ( 18 . 8% ) from four provinces had NAb titers ranging from 32× to 128× ( Fig 4 ) , of which five were both WB and ELISA positive , with the other four negative for both . The positive control ( H anti-SEOV ) had an NAb titer of 513× ( Fig 4 ) . Of interest is that one serum from Rousettus leschenaultii bat in Xishuangbanna , BN78 , had the highest NAb titer ( 128× ) against SEOV and was WB and ELISA positive only for SEOV , not for LAIV or XSV . Results also showed that some WB and ELISA double-positive bat sera against the three viruses did not neutralize SEOV , including samples BN19 , CZ63 , CZ26 , NP39 and ZJ61 .
As bat-borne HVs have only recently been identified , there is insufficient sequence data at present to provide a comprehensive analysis of their genetic diversities . Apart from the four complete genomic sequences reported here ( Table 2 ) , the sequences of representative bat-borne HVs published to date in GenBank show that many are not of complete genomes or even of a full-length gene segment [5–7 , 11] . Currently Laibin , Longquan and Quezon viruses are the only bat-borne HVs approved by ICTV so far [6 , 9 , 12] . Among completed genomes , there are the complete coding sequences of the three gene segments of BRNV from the Czech Republic [13] . Here we report the first genomic sequence of XSV and show that XSV and BRNV should be considered new HV species awaiting ICTV consideration . Fortunately , there are partial L gene sequences available for all bat-borne HVs , which allowed construction of a phylogenetic tree ( using 314 bp ) , permitting comparison with rodent- and insectivore-borne HVs ( S5 Fig ) . This showed that current bat-borne HVs can be classified into nine species as listed in Table 2 , with almost every one having a specific bat genus as host . Of them XSV is the most notable bat-borne HV , which has been found in different locations in present and previous studies [7 , 8] , and showing significant nt variation although aa sequences of the genomic segments are conserved . The nt identity between currently identified XSV variants ranges from 76 . 1%-93 . 4% ( Table 2 ) . Furthermore the NP of XSV showed cross antigenicity with both SEOV and LAIV but the NP of LAIV showed no cross antigenicity with SEOV ( Fig 2B ) , indicating that XSV is antigenically closer than LAIV to SEOV and therefore an ideal focus for gaining insight into the role of bat-borne HVs in public health . Meanwhile the closer relationship of insectivore-borne HVs NVAV and ALTV to bat-borne than to insectivore-borne HVs in the NP and GP trees ( Fig 1 ) indicated that HVs from bats and insectivores could share the common ancestry for evolution [18 , 46] . In general our sequence comparisons and phylogenetic analyses show that bat-borne HVs had broad genetic diversities and had evolved worldwide within an independent and diverse phylogroup . In this regard , more extensive studies obtaining more complete sequences in extended areas will undoubtedly identify more novel bat-borne HVs in future . Although nine bat-borne HVs have been identified worldwide , the virus detection rate is low and in limited locations [5 , 7 , 9 , 11] . The present study investigated bats in 22 cities , but the viral RNA was found in only three bat colonies in three cities , and RNA-positive rates were only 1 . 4% ( 1/74 ) for LAIV in BS , 3 . 0% ( 5/168 ) for XSV in LB and 7 . 5% ( 3/40 ) for XSV in PE ( see Fig 5 and Table 1 ) . In contrast , seropositive rates are higher: 28 . 1% ( 9/32 ) for LAIV and 40% ( 2/5 ) for XSV in BS and LB respectively ( sera were not collected in PE ) ( Table 1 ) . Low viral RNA detection rates have also been reported in previous publications with 5 . 6% ( 1/18 ) for MGBV in Sierra Leone [5] , 3 . 1% ( 1/32 ) for LAIV BT20 in China [9] , and 0 . 8% ( 1/123 ) for MAKV in Gabon [11] . Unfortunately , the seroprevalence was not reported in these publications . The antibody titers of bat sera against HVs in our study were rather low ( the most were 100× and only 18 were 400× ) as compared to those of rodent reservoirs which were usually higher and could reach 50 , 000 [24 , 47] , this might be ascribed to the higher diversity of VH ( especially FR3 ) in immunoglobulin genes of bats in comparison to those of mouse , swine and human [48] . The correlation between RT-PCR positivity and antibody positivity about hantavirus infection in rodents or shrews were reported [49 , 50] . Song et al . reported that a certain proportion , although not all , of Ussuri white-toothed shrews ( Crocidura lasiura ) with IgG antibodies against Imjin virus ( MJNV , a newly isolated hantavirus ) had MJNV RNA detectable by RT-PCR [50] . In our study bats sampled in 2015 and 2016 showed higher seropositive rates , but HV RNA was not detected from either seropositive or seronegative bats . All nine RT-PCR positive samples were collected in between 2012–2014 , but their antibody titers were not tested since the sera were not collected during that time . Moreover , serology study on bat-borne HVs was not conducted in previous publications , therefore further study is needed to understand the dynamics of HV infection and its antibody response in bats . Serological epidemiology is important to uncover the real situation of bat-borne HV prevalence , and is critical for eventual estimation of the potential risk of these viruses to public health . Since bat-borne HVs have never been isolated , their NP is a preferential target for serological investigation and antigenic differentiation . It is the main immunogenic protein which contains both serotype-specific and group common epitopes , and is commonly used as diagnostic antigen for HV detection [22–25 , 51] . For these reasons , rNPs of LAIV , XSV and SEOV expressed in E . coli were used to assay all 709 bat serum samples by ELISA , resulting in identification of a large number of seropositive sera ( Fig 3 ) , with many likely to cross react with two or three rNPs ( see OD492 values in S2 Fig ) . To confirm this , 88 ELISA sera were further tested by WB against all three rNPs with results showing that , except for sera reacting exclusively with one rNP , some could cross recognize two rNPs , mainly the rNPs of LAIV/XSV , or XSV/SEOV , and seldomly the rNPs of LAIV/SEOV ( see S3 Fig ) . It is notable that 21 sera could cross recognize three rNPs with 10 showing very strong reactivity against all three rNPs . The role of NP in producing this cross antigenicity was further verified by WB using a combination of eukaryotically expressed rNPs and NP-specific antiserum ( see Fig 2B ) . To identify NAb against SEOV , 48 bat sera were analyzed by FAVNT , which identified 9 ( 18 . 8% ) positives . Fig 4 summarizes the results of 48 bat sera assayed by FAVNT , ELISA and WB . Of the nine NAb-positive sera , four ( BN5 , ZS7 , ZS25 , ZS27 ) neutralized but did not react by ELISA or WB , three ( BN64 , CZ67 , NP6 ) not only neutralized but also reacted with three rNPs by ELISA and WB . The most interesting bat serum was BN78 , which neutralized SEOV and reacted with the rNP of SEOV but not with that of LAIV and XSV . BN78 was collected from a Rousettus leschenaultii bat , of this species 51 of 142 individuals showed anti-HV antibody positive ( 35 . 9% ) , the highest among all bat species ( Table 1 ) . Furthermore another Rousettus species ( Rousettus amplexicaudatus ) was reported to harbor Quezon virus in the Philippines [12] , suggesting that fruit bats in genus Rousettus are likely major reservoirs of HVs . Moreover many sera without neutralizing activity reacted with the three rNPs by ELISA and WB . Altogether , it is interesting to have found multiple patterns of cross-reactivity with three rNPs . Illustration of the complex patterns will be difficult but likely to imply that the bats had been infected with other unknown HVs . The prime example is bat serum BN78 . It had the highest neutralization titer against SEOV and exclusively strong reactivity with the rNP of SEOV , indicating that this bat was infected by an unknown HV antigenically very close to SEOV , but not SEOV since both human anti-SEOV convalescent serum ( Fig 2A ) and SEOV-specific anti-rNP serum ( Fig 2B ) could also cross react with the rNP of XSV . Altogether , the multiple genetic diversities and different cross-reactivity patterns indicate that more as yet unknown bat-borne HVs circulate in the investigated region , but to uncover them further investigation is needed . Viruses usually have a defined host range for circulation in nature . It is interesting to note that LAIV BT33 in the present study and LAIV BT20 in a previous study [9] have been identified in different locations but from the same bat species ( Ta . melanopogon ) . All XSV variants in the present and previous studies [7 , 8] have come from a single bat species , Hipposideros ( Hi . pomona ) . Apart from MAKV in Gabon , which also came from a Hipposideros sp . ( Hi . ruber ) , all other six HVs came from bats in six different genera . The host range of currently identified bat-borne HVs is summarized in Table 2 . In general , bat-borne HVs have a huge genetic diversity with different viral species harbored by different genera of bats , showing wide range of hosts . But regarding a given bat-borne HV species its host range may be narrow , restricted mainly to one bat genus . This implies that a given bat-borne HV may have a host tropism . As shown in Fig 5 , 12 of the 13 cities in 4 provinces in which serum collections were made had a positive seroprevalence , with Guangxi having the most positive samples and most seropositive locations ( 6 of 7 sampled cities were seropositive ) . Furthermore , two HVs were detected in two of its cities , BS and LB . In 2015 , LAIV was identified in LB [9] , in which XSV was found in the present study although from another location within the city , indicating that divergent bat-borne HVs co-exist in LB . LAIV was also found in BS this time , several hundred kilometers west of LB ( see Fig 5 ) , indicating that LAIV has a broad distribution in Guangxi province . It is notable that XSV has been identified in two north Vietnamese provinces , Tuyên Quang ( TQ ) and Phú Thọ ( PT ) , as shown in Fig 5 , and in the central Vietnamese province Quảng Nam since 2013 [7 , 8] . In present study eight strains of XSVs were identified in LB of Guangxi and PE of Yunnan , indicating that XSV circulates in the vast area between Chinese Guangxi/Yunnan and Vietnam . The accumulated serological and molecular data highly support the proposition that a vast area between China and southeast Asia provides a natural focus for bat-borne HV circulation . In this area natural circulation of genetically divergent bat-borne HVs in their hosts would be maintained , and therefore the concept of a bat-borne HV circulation sphere has been introduced to describe the situation . While there is a lack of sufficient serological data in Yunnan Province , a narrow area between southwest Guangxi and north Vietnam likely forms a main circulation sphere of at least two species of bat-borne HVs ( Fig 5 ) . With more extensive investigations this area may be extended , particularly to surrounding areas in Laos , Myanmar and even Thailand . In conclusion , the present study has compiled the first profiling of cross antigenicity between bat-borne and human-infecting HVs as well as among bat-borne HVs . It has also revealed the seroprevalence and wide distribution of bat-borne HVs in south and southwest China . A comprehensive analysis based on genetic diversity , seroprevalence , cross antigenicity and host range of the viruses has helped identify an area between China and Vietnam as a main circulation sphere where at least two bat-borne HVs circulate in the bat population . Given the existence of bat-borne HVs genetically and antigenically close to human-infecting HVs , extensive studies should be emphasized in future to assess the potential risk of bat-borne HVs to public health .
The procedures for sampling bats and mouse experiments in this study were reviewed and approved by the Administrative Committee on Animal Welfare of the Institute of Military Veterinary Medicine ( Laboratory Animal Care and Use Committee Authorization permit JSY-DW-2010-02 for bats and JSY-DW-2015-02 for mice ) . All live animals were maintained and handled according to the Principles and Guidelines for Laboratory Animal Medicine ( 2006 ) , Ministry of Science and Technology , China . The 1 , 419 tissues ( lungs and intestines ) and 709 sera of 1 , 561 bats used in present study were archived and sub-packed samples stored at -80ºC following collection between 2012 and 2016 in 22 cities of Yunnan , Guangdong , Fujian , Zhejiang and Guangxi provinces , China , and used to investigate viruses than hantaviruses in our previous studies [32 , 38 , 52 , 53] . Bat species were identified morphologically and then molecularly by sequencing the bat mitochondrial cytochrome b gene from muscle tissue [54] . These bats were classified as belonging to 22 species within 9 genera and 6 families: Hipposideridae ( n = 598 ) , Rhinolophidae ( n = 358 ) , Vespertilionidae ( n = 172 ) , Miniopteridae ( n = 129 ) , Emballonuridae ( n = 74 ) and Pteropodidae ( n = 230 ) . Detailed sample information is shown in Fig 5 and Table 1 . The tissue samples were homogenized and subjected to RNA extraction using the RNeasy Mini Kit ( QIAGEN ) , and the RNA was reversely transcribed into cDNA which was processed for viral metagenomic analysis as described previously [55] . Serum samples were used in serological analyses . Identified HV-like contigs were subjected to BLASTn and BLASTx search ( https://blast . ncbi . nlm . nih . gov/Blast . cgi ) . The genomic positions of the contigs were decided using Hantaan virus strain 76–118 as the reference . RNA was extracted from bat tissues ( intestines with contents and lungs ) and screened by RT-PCR using pan-HVs nested primers targeting a 396-bp sequence of the conserved L segment [40] . Positive amplicons were sent for Sanger sequencing ( Comate ) and the obtained sequences were used to initially determine their phylogenetic locations . Details of the primers can be seen in S2 Table . HV positive lung tissues ( ~100 mg/bat ) were thoroughly ground with DMEM ( 500 μL/100 mg tissue ) in a homogenizer and clarified by centrifugation at 5 , 000 g for 5 min . Following sterilization by passage through a 0 . 22 μm Millipore filter supernatants were incubated with BHK-21 cells as well as the HV-sensitive cell lines Vero and Vero E6 [56] ( all stored in our laboratory ) in 24-well plates . After incubation for 24 h at 37ºC , the cells were washed 2x with PBS , incubated in DMEM with 2% FBS for 14–21 days and observed daily for cytopathic effects ( CPE ) . Cultures , if showing no CPE , were harvested by freeze-thawing 3x and passaged again in the same cell lines . After five passages , the cultures were analyzed for HV by RT-PCR . To characterize the full genomic sequence and structure of detected HVs , genome-amplifying overlapping primers were synthesized based on the contigs ( S1 Table ) and representative sequences of previously published HVs including bat-borne HV LAIV BT20 [9] . Since the terminal nucleotide sequences of S , M and L segment are conserved among members of the genus Orthohantavirus [1] , their sequences were used as primers to obtain the end sequences of each segment . The targeting amplicons were amplified using the Phusion High-Fidelity PCR Master Mix with HF Buffer ( NEB ) with the recommend reaction system , and cloned into a blunt-end pLB vector using a Lethal Based Fast Cloning Kit ( Tiangen ) . Three clones of each amplicon were further identified by PCR and then sent for commercial Sanger sequencing . The complete genomic sequences were obtained by assembling amplicons with overlapped regions using SeqMan in the DNAStar software package . ORFs of each segment were searched by ORFfinder ( https://www . ncbi . nlm . nih . gov/orffinder/ ) in NCBI and the predicted proteins were further confirmed by aligning in BLASTp . The representative sequences of each classified HV species as well as some unclassified HVs were downloaded from GenBank . Their complete NP , GP and RdRp coding regions ( aa ) were aligned with those obtained in the present study using the online program MAFFT version 7 ( https://mafft . cbrc . jp/alignment/server/ ) . The best-fit substitution model for each tree was selected based on the Akaike Information Criterion ( AIC ) of Smart Model Selection ( SMS ) in PhyML ( version 3 . 0 ) [57] . Phylogenetic trees , including their topology and support for tree nodes , were then inferred using the maximum likelihood method , Subtree Pruning and Regrafting ( SPR ) , approximate Likelihood Ratio Test ( aLRT ) with the Shimodaira-Hasegawa-like ( SH-like ) tree-selection method in PhyML [58] . Sequence identities were calculated by MegAlign in DNAStar software package . The complete NP coding sequences ( CDS ) of bat-borne HV LAIV , XSV strains identified in the present study were amplified with the 5’ EcoRI and a 3’ XhoI sites at the two ends . The complete CDS of SEOV strain GuangzhouRn36 ( 1 , 287 nt , Accession number: GU592948 ) was optimized using E . coli codons and chemically synthesized ( GENEWIZ ) with the same restriction enzyme sites ( primers shown in S2 Table ) . The NP gene fragments so obtained were subcloned into a prokaryotic expression vector pET-28a ( + ) with a His-Tag at C terminus and used to transfected E . coli strain Rosetta ( Tiangen ) . The rNPs were expressed after induction with 0 . 5 mM isopropyl-β-d-thiogalactoside ( IPTG ) and identified by SDS-PAGE and WB using mouse anti-6X His-tag monoclonal antibody and Alexa Fluor 680-conjugated donkey anti-mouse IgG H&L ( Abcam ) as the respective primary and secondary antibodies . The rNPs of the three HVs were purified and quantified by Ni-NTA His Bind Resin ( Novagen ) and BCA Protein Assay Kit ( CWBio ) , following which they were identified using an SEOV-convalescent human serum ( H anti-SEOV ) and a negative human serum ( H Neg . ) as controls ( both stored in our laboratory ) . The H anti-SEOV serum was collected from a SEOV-infected convalescent patient who was diagnosed at onset by clinical symptoms and RT-PCR . Specific hyperimmune sera were prepared by injecting four week-old female Kunming mice intramuscularly with purified rNPs of the three HVs . Each injection contained 20 μg protein mixed 1:1 ( V/V ) with the Quick Antibody-Mouse 5W adjuvant ( Biodragon ) as recommended by the producers and booster doses with the same formulated rNP were given at 14 days later . At 21 days post boost , blood was collected through heart puncture for serum preparation . Antibody titers were determined by rNP-based ELISA as described below . Complete CDS of NPs of the same LAIV , XSV and SEOV strains were amplified with primers containing a 5’ XhoI site and a 3’ EcoRI site . Three NP fragments were fused in-frame to the C-terminal of the enhanced green fluorescent protein ( EGFP ) tag ( 239 aa , MW 27 kDa ) of eukaryotic expression vector pEGFP-C1 . Transient expressions of rNPs were obtained by transfection of BHK-21 cells in 6-well cell plates with the constructed plasmids with the blank vector pEGFP-C1 as control , using Lipofectamine2000 reagent ( Invitrogen ) according to the manufacturer’s protocol . The transfected cells were cultured at 37ºC with 5% CO2 for 36 h , and then examined microscopically for the expression of fusion proteins with green fluorescence . NP-expressing cell cultures were collected and lysed with Cell Lysis Buffer ( CST ) , and the total protein was quantified using the BCA Protein Assay Kit . Correct expression of rNPs in cell lysates were confirmed by WB using mouse anti-EGFP monoclonal antibody ( Abcam ) and Alexa Fluor 680-conjugated donkey anti-mouse IgG H&L as the respective primary and secondary antibodies . Eukaryotically expressed NPs were used as antigens to detect antibodies in mouse hyperimmune sera by WB . Briefly the eukaryotically expressed NPs were separated by SDS-PAGE and transferred onto nitrocellulose blotting membranes ( GE Healthcare ) , blocked with 5% non-fat milk ( Promega ) at 4ºC overnight , then incubated with the above three mouse anti-NP hyperimmune sera at 1:300 dilution for 2 h . After washing with PBST 3x , the membranes were incubated with Alexa Fluor 680-conjugated donkey anti-mouse IgG H&L 1:1 , 000 for 50 min . After washing , the membranes were scanned and photographed using an Odyssey imager ( LI-COR Biosciences ) . All WB analyses were repeated at least three times . ELISAs using rNPs as coating antigen were developed to detect antibodies of bat against LAIV , XSV and SEOV . Briefly 96-well microplates ( Corning ) were coated with purified prokaryotically-expressed rNPs ( 200 ng/well in NaHCO3-Na2CO3 buffer at pH 9 . 6 ) at 4°C overnight and blocked with 5% non-fat milk ( Promega ) at 37°C for 1 h . Then 100-fold PBS-diluted serum samples were added to the wells ( duplicate wells per serum ) and incubated at 37°C for 1 h followed by addition of HRP-conjugated goat anti-bat IgG , H&L chain , polyclonal antibody ( BETHYL , react specifically with bat IgG , and with light chains common to other bat immunoglobulins ) at a 1:20 , 000 dilution for bat sera , and HRP-conjugated goat anti-human IgG polyclonal antibody ( Zsbio ) at a 1:200 dilution for human sera ( controls ) . After incubation at 37ºC for 50 min , freshly-prepared O-phenylenediamine ( OPD ) substrate solution ( Sigma ) was added to each well for 5 min for color reaction , which was stopped by addition of 2 M sulfuric acid . The OD492 values were immediately read and blanked by the OD630 value using a Multimode Microplate Reader ( Infinite 200 PRO , Tecan ) . The cut-off values of bat sera were determined based on the highest CR between results of ELISA and WB . To validate the ELISA result , WBs using prokaryotically expressed rNPs were performed to detect bat and human sera . Briefly , the rNPs were separated by SDS-PAGE , transferred onto nitrocellulose membranes followed by blocking using the protocol described above . Membranes were then incubated with 1:300 diluted selected bat or human serum for 2 h , followed by 2 , 000-fold diluted HRP-conjugated goat anti-bat IgG H&L chain polyclonal antibody or 100-fold diluted HRP-conjugated goat anti-human IgG polyclonal antibody for 50 min . All procedures were conducted at ambient temperatures . Bands were pictured by automatic chemiluminescence ( Tanon ) . To determine the NAb titers of bat and human sera , the FAVNT using 200 TCID50 SEOV ( 105 . 31TCID50/0 . 1mL ) in Vero E6 cells was performed by a previously published protocol [59] . Anti-HV monoclonal antibody provided by the Fourth Military Medical University [60] was labeled using FITC [61] . The NAb titer of each serum was calculated using the Spearman-Kärber formula [62] . To compare the differences of the OD492 values of bat and human sera to each HV , normal distribution tests were conducted separately , and multiple comparison was performed using t tests ( LSD ) , Student-Newman-Keuls ( SNK ) test , Tukey's studentized range ( HSD ) test , Bonferroni ( Dunn ) t tests and Scheffe's test . Then κ test , an inter-rater agreement statistic , was used to evaluate the consistency between the results of ELISA and WB . All statistics were programed and calculated using the Statistical Analysis System ( SAS ) version 9 . 2 . | Some HVs are life-threatening pathogens predominantly carried and transmitted by rodents . In recent years bat-borne HVs have been identified in a broad range of bat species . To understand their significance to public health the present study conducted extensive investigations on genetic diversity , seroprevalence , distribution and cross antigenicity of bat-borne HVs in south and southwest China . The results provide the first profiling of cross-reactivity between bat-borne and human-infecting HVs , demonstrating that some bat sera can neutralize SEOV in cell culture . They also revealed that divergent bat-borne HVs co-exist and are widely distributed in Chinese Guangxi/Yunnan as well as in north Vietnam , resulting in identification of an area between China and Vietnam in which natural circulation of bat-borne HVs is maintained . Given the existence of bat-borne HVs genetically and antigenically close to human-infecting HVs , the need for extensive future studies is emphasized in order to assess the potential risk of these viruses to public health . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"taxonomy",
"medicine",
"and",
"health",
"sciences",
"enzyme-linked",
"immunoassays",
"immune",
"physiology",
"pathology",
"and",
"laboratory",
"medicine",
"china",
"immunology",
"geographical",
"locations",
"vertebrates",
"animals",
"mammals",
"invertebrate",
"genomics",
"phylogenetics",
"data",
"management",
"phylogenetic",
"analysis",
"immunologic",
"techniques",
"cross",
"reactivity",
"research",
"and",
"analysis",
"methods",
"immune",
"system",
"proteins",
"computer",
"and",
"information",
"sciences",
"serology",
"proteins",
"antigens",
"immunoassays",
"evolutionary",
"systematics",
"animal",
"genomics",
"people",
"and",
"places",
"biochemistry",
"eukaryota",
"asia",
"physiology",
"bats",
"genetics",
"biology",
"and",
"life",
"sciences",
"genomics",
"evolutionary",
"biology",
"amniotes",
"organisms"
] | 2019 | Seroprevalence, cross antigenicity and circulation sphere of bat-borne hantaviruses revealed by serological and antigenic analyses |
Dengue transmission by the mosquito vector , Aedes aegypti , occurs indoors and outdoors during the day . Personal protection of individuals , particularly when outside , is challenging . Here we assess the efficacy and durability of different types of insecticide-treated clothing on laboratory-reared Ae . aegypti . Standardised World Health Organisation Pesticide Evaluation Scheme ( WHOPES ) cone tests and arm-in-cage assays were used to assess knockdown ( KD ) and mortality of Ae . aegypti tested against factory-treated fabric , home-dipped fabric and microencapsulated fabric . Based on the testing of these three different treatment types , the most protective was selected for further analysis using arm-in cage assays with the effect of washing , ultra-violet light , and ironing investigated using high pressure liquid chromatography . Efficacy varied between the microencapsulated and factory dipped fabrics in cone testing . Factory-dipped clothing showed the greatest effect on KD ( 3 min 38 . 1%; 1 hour 96 . 5% ) and mortality ( 97 . 1% ) with no significant difference between this and the factory dipped school uniforms . Factory-dipped clothing was therefore selected for further testing . Factory dipped clothing provided 59% ( 95% CI = 49 . 2%– 66 . 9% ) reduction in landing and a 100% reduction in biting in arm-in-cage tests . Washing duration and technique had a significant effect , with insecticidal longevity shown to be greater with machine washing ( LW50 = 33 . 4 ) compared to simulated hand washing ( LW50 = 17 . 6 ) . Ironing significantly reduced permethrin content after 1 week of simulated use , with a 96 . 7% decrease after 3 months although UV exposure did not reduce permethrin content within clothing significantly after 3 months simulated use . Permethrin-treated clothing may be a promising intervention in reducing dengue transmission . However , our findings also suggest that clothing may provide only short-term protection due to the effect of washing and ironing , highlighting the need for improved fabric treatment techniques .
Dengue is the “most important mosquito-borne viral disease” [1] with over 3 . 5 billion people at risk of infection [2] , 50–100 million new infections and 20 , 000 deaths annually [1] . Most control strategies target the main urban vector , Aedes aegypti [3] , but despite multiple control programmes designed to reduce the immature and adult stages of the mosquito [4] , Ae . aegypti has continued to increase its distribution over the past 25 years [5] . Since Ae . aegypti is a day-biting mosquito , technologies that are worn during the day may offer protection against mosquito bites and have the potential to reduce dengue transmission [6] . Personal protection technologies function by repelling or killing the vector , or by providing a physical barrier between vector and host [7 , 8] . The best known personal ( and community ) protection technology is long lasting insecticidal nets ( LLIN ) commonly used in malaria control programmes [9] . LLIN’s are most effective against mosquitoes that feed during the night , but surprisingly have been shown to provide some protection against day biting vectors such as Ae . aegypti presumably because they may rest on treated netting [10] . Insecticide-treated clothing is an intervention that could protect individuals during the day , when users are at work or school , and could easily integrate into everyday routines . Indeed , agricultural and wildlife groups , commercial companies and the military currently use insecticide treated clothing [11–13] . A recent modelling study demonstrated that insecticide-treated school uniforms can reduce dengue burden by up to 50% in school children in Thailand , highlighting the potential impact insecticide treated clothing could have on dengue transmission [14] . For the clothing to be a sustainable intervention it must be safe , effective and long-lasting . It must also be able to withstand regular washing , be low—cost and acceptable to members of the local communities . All these factors are influenced by the active ingredient and the type of treatment method . Permethrin , a commonly used pyrethroid , is the main compound used in treated clothing [15] . There are several techniques used currently for treating material with an insecticide . They include absorption , incorporation , polymer coating , and micro-encapsulation [16] . Despite the common use of permethrin-treated clothing in the military and recreational industries , studies have been published on its ability to knockdown ( KD ) , kill , repel and prevent biting [15] but they have not investigated variation between clothing treatment techniques . Other important undetermined factors that might affect efficacy of the treated clothing include repeated washing , exposure to Ultra-Violet ( UV ) light and heat exposure ( for example , caused by ironing ) . These factors could have a significant effect on the efficacy and duration of protection provided by impregnated clothing when used in the field . The aim of this study was to investigate the efficacy and duration of protection provided by permethrin-treated clothing which could be later used in a randomised controlled trial of permethrin-treated school uniforms to protect children from Dengue fever in Thailand [17] .
The following types of clothing were used in this study: Aedes aegypti , ( pyrethroid susceptible strain ) were obtained from reference strain ( originally from West Africa , colonised in 1926 with field additions in 1976 ) held at LSHTM , UK . All mosquitoes were reared and housed under optimal environmental conditions of 25°C ± 2°C and 80% RH with a 12: 12 hour photoperiod . Ages ranged for all testing from 3–7 days old ( 3–5 for cone tests and 5–7 for arm-in-cage ) . All mosquitoes used in experiments were nulliparous females fed on 10% glucose solution . This study was approved by the London School of Hygiene and Tropical Medicine Ethics committee ( reference number 6074 ) . The participant used in this study was aged between the ages of 18–65 and provided written informed consent before taking part in this study .
No difference in knockdown after 3 minutes was shown between factory-dipped clothing ( FDC ) , factory dipped school uniforms ( FDSU ) and microencapsulated clothing ( MC ) ( Table 1 ) . After 1 hour exposure , the FDC and FDSU produced a knockdown of 96 . 5% ( 95% CI 94 . 2%-97 . 8% ) and 98 . 2% ( 95% CI 92 . 1%-99 . 6% ) respectively ( Table 1 ) . MC produced a lower 1 hour knockdown , 50 . 8% ( 95% CI 37 . 5%-62 . 7% ) and 24 hour mortality , 73 . 3% ( 95% CI 51 . 8%-87 . 4% ) ( Table 1 ) when compared to both FDC ( 1 hour KD , p = 0 . 0043 , 24 hour mortality p = 0 . 0331 ) and FDSU ( 1 hour KD , p = 0 . 0089 and 24 hour mortality , p = 0 . 0384 ) . There was no difference between FDC and FDSU across any of the time points . There was a significant difference between mean number of mosquitoes landing between bare arm , control clothing and treated clothing ( p = 0 . 0001 ) . Control and treated clothing did provide significant protection against landing and biting when compared to the bare arm control ( p = 0 . 0001 ) , with the exception of the microencapsulated ( MC ) untreated control clothing where there was no significant difference between the control clothing and bare arm ( p = 0 . 5776 ) ( Fig 2 ) . Bite protection for MC was significantly lower at 65 . 5% ( 95% CI 57 . 7%-72 . 4% ) compared to HDC 91 . 5% ( 95% CI 80 . 9%- 96 . 3% ) , ( p<0 . 0001 ) , ( Fig 2 ) . MC gave a bite protection of 79 . 9% ( 95% CI 70 . 41%-86 . 32% ) and this is not significantly different when compared to FDC ( p = 0 . 538 ) . No significant difference was found between HDC and FDC for both landing and biting protection ( HDC bite protection: 91 . 5 , 95% CI 80 . 9%-96 . 3% ) ; HDC landing protection: 49 . 9% ( 95% CI 44 . 4%- 55 . 0% ) ; FDC bite protection: 79 . 9% ( 95% CI 70 . 41%- 86 . 32% ) ; FDC landing protection 40 . 9% ( 95% CI 36 . 4%-45 . 1% ) ( p = 0 . 999 ) , with both FDC and HDC providing the greatest protection against both landing and biting .
The clear reduction in the number of bites an individual receives , combined with the high mortality and knockdown caused by permethrin-treated clothing , is proof that insecticide-treated clothing could be a promising additional intervention for dengue prevention . It has the potential to reduce the number of Aedes mosquito bites thereby reducing disease transmission . However , for the clothing to be used successfully , improved methods of treatment are needed to ensure duration of protection provided is increased and cost-effectiveness is achieved . A study evaluating field-like conditions would be beneficial to better understand the effect of washing and environmental exposure under natural conditions . In addition , the protection provided by permethrin treated clothing when wearing partial coverage clothing ( i . e . shorts and short sleeved shirt ) and resistant mosquito populations should be performed . | Personal protection technologies could be a key tool in the fight against arthropod borne diseases . Insecticide treated clothing may have a significant effect on reducing mosquito borne disease by reducing biting rates and local vector populations . Currently there are four main treatment techniques; absorption , incorporation , polymer coating , and micro-encapsulation which are used to treat clothing with permethrin but little is known regarding the efficacy and duration of protection provided by these techniques . The evaluation of these different treatment techniques using standardised World Health Organisation Pesticide Evaluation Scheme ( WHOPES ) assays and high pressure liquid chromatography ( HPLC ) analysis provides further information on factors which have a significant effect on the efficacy and duration of protection of treated clothing . This will aid in the design and implementation of control programs using insecticide treated clothing . However , washing technique and heat exposure have a significant effect on efficacy , emphasising the need for further investigation into treatment techniques , so duration of protection can be increased . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [] | 2015 | Permethrin-Treated Clothing as Protection against the Dengue Vector, Aedes aegypti: Extent and Duration of Protection |
Robustness , defined as tolerance to perturbations such as mutations and environmental fluctuations , is pervasive in biological systems . However , robustness often coexists with its counterpart , evolvability—the ability of perturbations to generate new phenotypes . Previous models of gene regulatory network evolution have shown that robustness evolves under stabilizing selection , but it is unclear how robustness and evolvability will emerge in common coevolutionary scenarios . We consider a two-species model of coevolution involving one host and one parasite population . By using two interacting species , key model parameters that determine the fitness landscapes become emergent properties of the model , avoiding the need to impose these parameters externally . In our study , parasites are modeled on species such as cuckoos where mimicry of the host phenotype confers high fitness to the parasite but lower fitness to the host . Here , frequent phenotype changes are favored as each population continually adapts to the other population . Sensitivity evolves at the network level such that point mutations can induce large phenotype changes . Crucially , the sensitive points of the network are broadly distributed throughout the network and continually relocate . Each time sensitive points in the network are mutated , new ones appear to take their place . We have therefore named this phenomenon “whack-a-mole” sensitivity , after a popular fun park game . We predict that this type of sensitivity will evolve under conditions of strong directional selection , an observation that helps interpret existing experimental evidence , for example , during the emergence of bacterial antibiotic resistance .
Robustness , defined as tolerance to perturbations such as mutations and environmental fluctuations , is pervasive in biological systems [1 , 2] . Early computational models of evolution aimed at understanding the relationship between gene-network evolution and behavior ( gene expression dynamics ) [3–5] . These studies found that , although a large number of different networks ( genotypes ) have the same gene expression dynamics ( phenotype ) , they can usually be connected to one another via minimal changes ( e . g . creation or deletion of single cis-regulatory interactions ) that might easily occur during evolution via mutation . This capacity for neutral evolution can facilitate the evolution of robustness since it allows a population to migrate towards more robust genotypes without altering the phenotype [6] . Numerous theoretical studies have shown that robustness will evolve in particular when the phenotype is under evolutionary pressure to remain constant ( stabilizing selection ) . Experimental results are consistent with this notion . Gene networks in E . coli , for example , have been shown to be robust specifically to regulatory rewiring [7] . Similar experiments on metabolic networks , also in E . coli , have shown network robustness with respect to both gene knockouts and network rewiring [8–10] . Many recent studies have shown that ecological interactions both within and between species , and particularly coevolutionary interactions , drive evolutionary changes on a far more rapid timescale than previously estimated [11–13] . Here we use network modeling to understand how coevolutionary selection , rather than stabilizing selection , evolves network structure and function and how coevolution determines evolutionary properties such as robustness and evolvability [11] . We focus on a simple case of antagonistic coevolution between two populations , specifically a parasite population that uses mimicry of a complex phenotype as its survival strategy , as well as its host population . There are many documented cases of such interactions . A well-studied example is brood parasitism of cuckoos on their avian hosts . For instance , cuckoo finches ( Anomalospiza imberbis ) deposit their eggs in the nest of their host , the African tawny-flanked Prinia ( Prinia subflava ) . By mimicking the eggshell morphology of their hosts , the cuckoos trick their hosts into brooding these eggs . An evolutionary arms race between cuckoos and their host species drives continued variation in eggshell morphology in both species [14 , 15] . In another example , coevolution of complex chemical signals occurs between Maculinea alcon , a parasitic butterfly species and their host , Myrmica ants [16] . M . alcon larvae emit a pattern of surface chemicals very similar to those of the ant larvae , leading the ants to adopt and feed the butterfly larvae as their own . An evolutionary arms race has arisen between these two species such that the ants evolve changes in their larval surface chemicals to discriminate their own larvae from those of the parasite whereas the parasite is continuously evolving to again produce a similar pattern . It has previously been suggested that evolvability—the capacity for generating new phenotypes—can be facilitated by robustness , a somewhat counter-intuitive idea since evolvability and robustness would superficially appear to be opposite concepts . However , mutations will tend to accumulate in populations with high robustness , leading to greater genetic variation , which in turn may facilitate access to new phenotypes [1 , 6] . Phenotypic variation might be accessible during episodes of directional selection or particular conditions such as environmental stress [1 , 17–20] . Thus , under this model , periods of stabilizing selection allow genetic variation to accumulate , which is then eliminated by periodic selective sweeps and the cycle begins again with a new period of stabilizing selection . At the same time , the importance of this model remains unclear since few studies of network evolution have gone beyond stabilizing selection to investigate more realistic selection regimes [21] . Here we analyze host-parasite coevolution and find an entirely different strategy arises in which networks evolve a capacity for evolvability together with robustness against mutations . Here , evolvability in the network facilitates coevolutionary adaptation and is distributed throughout the network . Previous studies have also shown there is a relationship between evolvability and modularity in networks . A strategy of using two target phenotypes presented , for example , in alternating succession has been used because it can select for distinct network modules , each of which is capable of generating one of the target phenotypes [22 , 23] . One such study by Kashtan and Alon [22] used feed-forward logic networks and found that modularity evolved together with a fixed “evolvability node” which controlled the switch between two modules when mutated , thus switching phenotypes . Subsequent analyses showed evidence for modularity in other contexts including neural and metabolic networks [22 , 24–26] . An alternative to a fixed “evolvability node” may be to have evolvability distributed throughout the network , allowing phenotype changes to occur in many different ways . Both types of evolvability are observed in nature [27] . Examples of fixed evolvability nodes include the Drosophila shavenbaby locus which predominantly controls trichome patterning [27] , Pitx1 which determines the pelvic spine phenotype in stickleback fish [28] and optix which controls rapidly evolving wing patterns in Heliconius butterflies [29] . Examples of distributed evolvability have been reported in bacterial and virus species including in Helicobacter pylori where a broad spectrum of genetic variations explains adaptation to its human host [30] , in the pathogen Pseudomonas aeroginosa where antibiotic resistance evolves via several different mechanisms [31] and similarly in E . coli adaptation to low glucose environments [32] . Although the examples above illustrate the two extremes of what is likely a continuum between fixed and distributed evolvability , here we investigate a more general question—what conditions might favor the evolution of fixed vs distributed evolvability ?
To study gene regulatory network evolution under antagonistic coevolution we defined a model with two interacting populations . The model is an extension of a widely used single-population model that assumes stabilizing selection [33 , 34] . As in the previous model , each population functions at two broad levels: genotype-to-phenotype mapping and population dynamics ( see Methods for details ) . For the genotype-phenotype mapping , the genotype is defined as a gene regulatory network of N genes represented by a N×N matrix , W , the entries wij of which represent the regulatory strength and sign of gene j on gene i ( N = 10 was used for all results unless otherwise stated ) . The genotype is mapped to phenotype via gene expression dynamics . The gene expression levels at time t are represented by S ( t ) , a length N vector S ( t ) = [s1 , s2 , … , sN] ( 0≤si≤1 , i = 1 , … , N ) . The genotype W defines a dynamical system that is used to determine steady state gene expression levels for each gene , which correspond to the phenotype , S^ . Both host and parasite populations have a fixed number of individuals M . Cycles of reproduction , mutation and selection proceed in parallel as shown schematically in Fig 1A . Reproduction ( either sexual or asexual ) and mutation largely follow previous models [35–37] . Genotype mutations allow for creation and deletion of regulatory interactions as well as quantitative changes [36] . The main difference with previous models is at the selection stage , where the host and parasite populations interact by mutually determining fitness in the other population . To represent antagonistic coevolution in our model we assume that a candidate parasite individual has higher fitness when its phenotype is similar to that of a randomly chosen host individual ( a new random host is chosen for each parasite at every selection step and similarly for each host ) . Thus parasite fitness is defined as: f ( S^P ) =e−D ( S^P , S^H ) α , where D ( X , Y ) =∑i=1N ( xi , yi ) 2N , S^P is the parasite phenotype , and S^H is the phenotype of a randomly selected host individual . On the other hand , we assume the host has higher fitness when its phenotype is different from that of the parasite and therefore host fitness is defined as: f ( S^H ) =e−1−D ( S^H , S^P ) α where S^P is the phenotype of a randomly selected parasite individual . α is a parameter representing selection pressure . The fitness functions are symmetric about x = 0 . 5 ( S1 Fig ) to avoid any bias in how selection is applied in host vs parasite . Although the initial phenotypes are random , this two-population approach allows the eventual target phenotypes to emerge from the model , in contrast to previous models where the target phenotypes are defined a priori . The fitness definitions used are analogous to the two examples of host-parasite evolutionary arms races described above ( cuckoo finch and M . alcon ) whereby similarity ( and differences ) in complex phenotypes are selected for: eggshell morphology in the case of the cuckoo finch or the pattern of larval surface chemicals in the case of M . alcon . Under sufficiently strong selection pressure ( α ) both host and parasite populations reach a stage where their phenotypes alternate between one phenotype S^ and an approximately “inverted” version of the same phenotype , i . e . 1−S^=[1−s1 , 1−s2 , … , 1−sN] . At a given generation , if the host population phenotype is S^H and that of the parasite is S^P=1−S^H , then the host will have high fitness and the parasite will have low fitness . However , if at a later generation the parasite population is able to “invert” its phenotype S^P→1−S^P ( =S^H ) and the host population maintains its phenotype ( S^H ) , then the parasite and the host phenotypes will become the same—the host will now have low fitness and the parasite will have high fitness . The parasite population will continue “winning” until the host population is able to invert its phenotype , and the cycle continues . Fig 1B and S2 Fig show this progression over time ( vertical axis ) for every gene expression level in every gene ( horizontal axis ) of every individual in a typical simulation ( see Methods for parameter values used ) . Each cell in Fig 1B is colored blue when the expression level favors the host “winning” ( i . e . when a host gene is on and the corresponding parasite gene is off and vice versa ) , and yellow if the parasite is “winning” ( i . e . the host and parasite levels are the same ) . We see that by generation ~100 ( blue arrow Fig 1B ) both populations have converged to an alternating strategy as the rows alternate in color . Thus both host and parasite genotypes have become highly evolvable in response to phenotype changes in the opposite population . We are primarily interested in how these coevolutionary interactions between host and parasite populations affect gene regulatory network evolution and in particular how evolvability itself evolves within the networks . As expected , under weaker selection ( approximately α>0 . 15 , see S3A Fig ) the alternating phenotype did not evolve , so we focused here on the stronger selection case . We next sought to identify the mechanism underlying the phenotype inversion process , i . e . the evolution of evolvability . One possibility is that the alternating phenotype strategy would evolve in the form of a particular “evolvability hotspot” or interactions in the network , analogous to those identified previously by Kashtan et al . [22] in modular networks . A mutation in an “evolvability hotspot” would be highly likely to cause a phenotype inversion . An alternative scenario is one in which the capacity for phenotype inversion is highly distributed , and phenotype inversion can occur in many different places throughout the network , albeit with low probability . To assess these effects we implemented two measurements: first , a sensitivity score ( SS ) that estimates the overall probability that a mutation will cause a phenotype inversion ( see Methods ) , and secondly a measure of how distributed the sensitivity is within the set of network interactions that can cause a phenotype inversion , as described below . In addition , to measure the effects of coevolution on the remaining parts of the network ( that do not cause phenotype inversions ) we also quantify mutational robustness in this subset of network interactions . Fig 2A shows the progression of the sensitivity score ( SS ) during a typical simulation . Here we see that at the beginning of the coevolutionary process , because host and parasite networks are random , both have a negligible number of sensitive interactions and the mean SS is close to zero . As antagonistic coevolution proceeds and both populations evolve towards the alternating phenotype strategy , they both acquire sensitive interactions and the mean SS increases , eventually reaching a plateau . For the set of parameters shown in this example ( see Methods ) , SS reaches approximately 0 . 08 . Although this qualitative behavior is observed across a large range of parameter values , there are quantitative differences . Thus , the steady state SS level is reduced , as expected , if selection pressure is lower ( S3A Fig ) and with smaller population sizes ( S3C Fig ) where random drift effects are greater . Also , networks with a greater density of connections can evolve sensitivity more easily ( S3E Fig ) . Lastly , note that multiple simultaneous mutations can occur within a single genotype , particularly when the frequency of the single mutation is high , as is often the case when a population is undergoing a phenotype inversion ( S4A and S4B Fig ) . Although such events occur at low frequency , we found that , in cases of double mutations at least one of the mutation positions had a high sensitivity score whereas the other usually had a sensitivity score that was either very low or zero ( S4C Fig ) . Thus , a phenotype inversion is most often achieved with a single point mutation at a sensitive interaction , although occasional double mutations where at least one mutation is at a sensitive interaction can also cause a phenotype inversion . The sensitivity score ( SS ) estimates the probability of causing a phenotype inversion . We found that the capacity for causing a phenotype inversion is distributed across a large number of sensitive network interactions , and we therefore sought to quantify how sensitivity was distributed throughout the network . Sensitivity might either be distributed fairly equally among these interactions , or unequally in the sense that particular interactions are likely to cause a phenotype inversion whereas others interactions do so only with low probability . To quantify the distribution of sensitivity we first chose the subset of network interactions , wij , that exhibit sensitivity , where the subset is defined as those having a ( interaction specific ) sensitivity score SSij>0 ( see Methods ) . We compared the observed standard deviation ( SD ) of the SSij values to the SD of a null model that assumes the observed total sensitivity in this set of nodes is randomly distributed ( see Methods ) . We consistently found that the null model has a comparable , and even slightly higher , variance of sensitivity within the sensitive interactions than the evolved networks ( S5A Fig ) . Thus the levels of sensitivity are at least as similar amongst themselves than would be expected by chance given the observed total sensitivity in the network ( see Methods ) . Apart from causing a phenotype inversion , a mutation may either ( a ) leave the phenotype unchanged , which indicates robustness , or ( b ) cause the phenotype to change only partially which will usually be sub-optimal . As a measure of robustness , Fig 2B shows the fraction of mutations that leave the phenotype unchanged if we exclude phenotype inversions , i . e . ( a ) / ( ( a ) + ( b ) ) . These results show that robustness initially decreases but then increases , eventually reaching a level higher than that of the initial population . Note that the initial host and parasite populations have random phenotypes , generally their phenotypes are not in either similar or inverted forms . In addition sensitivity does not exist before coevolution . Therefore during the initial phase , all host/parasite individuals are under evolutionary pressure to explore alternative phenotypes to counter the other ( parasite/host ) population , which is also in a similar situation . Partial phenotype changes will therefore be beneficial until both populations enter the process of phenotype inversion . This is why robustness decreases in the earliest stages of coevolution ( Fig 2B ) . However , once the capacity for phenotype inversion has evolved , partial phenotype changes will not be beneficial especially under strong selection and there is selection pressure for mutations to either preserve or invert the phenotype . This is why robustness increases together with sensitivity , and why robustness eventually exceeds the initial ( pre-selection ) levels . Again , we found that the phenomenon of increased robustness is observable across a wide range of parameter values although the range is more limited under sexual reproduction than it is with asexual reproduction ( S6 Fig ) . Generally though , robustness evolves in the parts of the network that are not causing phenotype inversion . Thus in the steady state , both robustness and evolvability coexist in the network under coevolutionary selection . Although we observe that mutational robustness evolves under antagonistic coevolution , environmental robustness appears to coevolve to a much lesser extent . Previous studies have shown that even without direct selection for environmental robustness , mutational and environmental robustness will coevolve under stabilizing selection [38 , 39] . Environmental robustness was evaluated via perturbations of the initial gene expression levels and then by measuring the phenotypic distance between the perturbed and unperturbed cases ( see Methods ) . Given the trend for mutational robustness ( Fig 2B ) , the overall pattern was similar to that expected ( S7 Fig ) . However , the phenotypic distance increased to steady state levels that were well above those observed initially , indicating an overall reduction in environmental robustness . This was the case regardless of whether the perturbation rates were low or high relative to the mutation rate . We have addressed the simple case of equal population sizes for host and parasite . This case is relevant to many real host-parasite interactions such as the example of the cuckoos and their avian hosts discussed above , where the populations appear to be relatively stable and of comparable size [40] . Clearly however , host and parasite populations will often differ in size . We therefore evaluated the case of host population size = 100 and parasite population size = 1000 ( and vice-versa ) , finding only slight differences with the case of equal population sizes ( S8 Fig vs . Fig 2 ) . However , due to computational constraints we were unable to model much larger population sizes and we therefore leave a more thorough evaluation of unequal population sizes for future work . We next investigated whether sensitivity is preserved at particular points in the network or whether it changes over time . As described above for the case of modular networks , sensitivity will often evolve to be focused on “hotspots” that control distinct phenotypes and which do not change over time [22 , 24] . To assess the changes in the sensitive interactions over time we used asexual reproduction . Under asexual reproduction , tracing the ancestral lineage is straightforward because there is a single parent for each individual , and after G generations each individual in the population needs G ancestral genotypes to store its genetic history . In contrast , under sexual reproduction each individual needs at most 21+22+⋅⋅⋅+2G ancestral genotypes , which rapidly becomes unwieldy . We consider the set of sensitive points of the network ( i . e . those interactions wij with sensitivity score SSij>0 that may cause a phenotype inversion ) and how this set changes over time . We selected networks at a particular steady state generation and compared these to ancestral networks at various evolutionary distances . The comparison was done by measuring the similarity , in terms of sensitivity , between the ancestral and derived networks using the Jaccard index ( see Methods ) as shown in Fig 3A . Given that the phenotype is constantly changing , to ensure a valid comparison we only compared with ancestral networks having the same phenotype . As shown in Fig 3A , the overlap in sensitivity remains high only for a short time period , before dropping almost to levels that would be expected by chance ( null model Fig 3A—also see Methods ) . However , at steady state the sensitivity remains stable , as do the total number of sensitive interactions ( generations ~1000 onwards , Fig 2A and S5B Fig ) . Thus , sensitive interactions are highly labile and on average , each time a sensitive interaction is eliminated by mutation , a new one emerges to take its place . Colloquially this property is known as “whack-a-mole” , named after the fun park game , and we therefore refer to this phenomenon as whack-a-mole sensitivity . Even though sensitive interactions are labile and are constantly being relocated , we thought there might be a specific subset of interactions with consistently high sensitivity . Alternatively there might be no persistence in the sensitive interactions or any such interactions would be rapidly lost . Consistent with the latter scenario we found there are no interactions with a significantly high frequency of being a persistent sensitive interaction within a population and throughout a simulation , as shown in Fig 4 . Fig 4A shows , for a typical simulation , the frequency at which each interaction wij was sensitive over a period of 1500 generations while sensitivity and robustness were at steady state levels . Fig 4B shows the change in sensitivity over time for two particular interactions in Fig 4A ( those that had the highest and lowest overall sensitivity respectively ) . Fig 4C shows the same data in histogram form ( green curve ) together with the mean value for many simulations ( red curve ) . Even though there appears to be no preference for particular positions within the matrix , we tested whether there was a higher-level preference for particular rows of the interaction matrix W , which represent the cis-regulatory elements for each gene . For this , we considered the total sensitivity score for each row ( i ) , SSi , and in particular , tracked the row imax for which the value of SSi is maximal within each individual ( S9A Fig ) . We found that rarely does a particular imax dominate both the population and throughout generations ( S9B Fig ) . We repeated this analysis for columns , which represent gene outputs regulating genes , finding similar results . Thus , there does not appear to be any predilection for sensitivity to be associated with particular genes . As explained in the introduction , another way by which phenotypic innovation has been proposed to occur is through increased genetic variation , which is promoted by robustness . However , if a sensitivity mechanism has evolved to generate the appropriate phenotype changes , it does not , in principle , require high levels of genetic variation to function . To investigate the observed levels of genetic variation in the population that has evolved sensitivity we used a measure that simplifies each network using the sign of each matrix entry sgn ( wij ) , then counts the number of distinct ( simplified ) networks , expressed as a fraction of the population . Fig 3B shows how this diversity measure increases over time . Taking a typical host case ( green curve ) as an example we found that in the final population there were 91 distinct networks , which expressed as a fraction of the total population , leads to a diversity measure of 91/200 = 0 . 45 . The average trend ( red and blue curves ) shows diversity increasing over time , eventually reaching a plateau . This diversity is a consequence of the beneficial mutations ( occurring at sensitive interactions ) being broadly distributed throughout the network , thus making multiple evolutionary pathways available . Taking this analysis further , we used the same diversity metric to measure the level of variation generated by stabilizing selection ( see generations 1–500 in S10 Fig ) and found that the level of diversity was consistently below that observed under antagonistic coevolution . Although the comparison needs to be interpreted cautiously given that stabilizing and coevolutionary selection are quite different , we include it here to emphasize the high degree of diversity observed under coevolution . This high diversity occurs despite there being , in principle , no requirement for it . Lastly , we assessed the impact of having an initial phase of stabilizing selection that allows each population to evolve robustness and accumulate genetic variation independently before the coevolutionary process begins . As shown in S10 Fig ( generation 500 onwards ) , genetic variation increases under the initial stabilizing selection phase ( previous model of [34] was used for this ) , reaching a plateau by generation ~100 . Once coevolution begins , at generation 500 , genetic diversity is reduced as both populations pass through a bottleneck but then increases eventually exceeding the level achieved under stabilizing selection . However , the dynamics are not significantly different from those observed without the initial phase of stabilizing selection , and therefore the initial phase appears not to offer any advantage . As mentioned in the Introduction , previous studies [22 , 24 , 26 , 38] have investigated the conditions leading to increased modularity in a network using multiple target phenotypes . In each case , the target phenotypes contained a combination of features such that a modular network evolves . In particular , the study by Espinosa-Soto and Wagner [23] used two target phenotypes that only overlapped partially , leading to increased modularity ( S11A Fig , red curve ) . Applying the same modularity measure to our own simulations ( see Methods ) , we found that increased modularity did not evolve for any combination of parameters . We thought this might be because in our model , the entire phenotype alternates , in contrast to only part of the phenotype in the Espinosa-Soto model . However , a variant of our model in which only half the phenotype genes participate in host-parasite fitness also did not evolve modularity ( S11A Fig , green curve ) . The other key differences between the two models are how gene interactions and expression levels are represented ( real vs discrete ) , the method of presentation of target phenotypes ( sequentially alternating vs simultaneous ) and the type of perturbation ( mutational vs environmental ) , as summarized in S11B Fig . We therefore tested variant models that contained mixtures of features from either of the models but were unable to find increased modularity for any of the variant models ( S11A Fig ) . These results suggest that modularity will evolve only under the very specific conditions . Biologically , the most important of these conditions is perhaps the nature of the perturbations , which can broadly be interpreted as growth-related or developmental for our model vs physiological or environmental in the case of the Espinosa-Soto model . Another relevant study , by Kashtan and Alon [22] also found that modularity evolved in the network together with persistent sensitive nodes . This was achieved using alternating target outputs , so-called Modularly Varying Goals ( MVGs ) , which were defined as pairs of logical functions containing different combinations of sub-goals . For example , the authors defined two functions ( of 4 inputs , X , Y , Z and W ) as G1: ( X XOR Y ) OR ( Z XOR W ) , G2: ( X XOR Y ) AND ( Z XOR W ) . Although the model used was based on logical circuits and therefore quite different to the one we have used here , we evaluated whether using this particular pair ( G1 , G2 ) of MVGs would also result in long-term sensitive nodes . We implemented this using a single population model with networks having 4 designated input genes and 6 interacting regulatory genes , one of which is considered the output . Since there are 24 = 16 possible inputs , fitness was defined as the fraction of correct input-output mappings . We evolved the population using alternating targets ( G1 , G2 , G1 , … ) for 50 generations per target . For a population evolved under one target ( e . g . , G1 ) , we assessed sensitivity , and in particular we considered any mutated network as sensitive if it matched the alternate target ( e . g . , G2 ) in more than 12 out of 16 input-output pairs ( i . e . , a fraction of 0 . 75 ) . The threshold was set to 0 . 75 because we did not observe the average fitness exceeding this level for either target ( S12A Fig ) . As shown in the S12B Fig , we do observe that a subset of persistent sensitive network nodes evolves; this is the subset of nodes with frequency of sensitivity equal to 1 . However , in contrast to the previous study we did not observe increased modularity over time ( S12C Fig ) , presumably because most sensitive interactions are not persistent , but highly transient , with frequencies of sensitivity between zero and one ( S12B Fig ) . To further confirm that it is indeed the MVGs that facilitate the appearance of the subset of persistent sensitive nodes , we checked two further scenarios using the single population model . Firstly , we used two alternating targets in which half of the target genes ( N/2 ) are kept the same as the founder phenotype and the other half are inverted every 50 generations . This model is similar to that of Espinosa-Soto described above , except that the targets alternate in time , rather than being selected for simultaneously . In the second case , we simply alternated between the founder phenotype and its inverted form , again every 50 generations . In neither of these cases did we observe the emergence of persistent sensitive nodes ( S13 Fig ) as we observed with the MVGs . A second important difference between our approach and that of Kashtan and Alon lies in the mutation model . The Kashtan and Alon study used only topology changes , whereas our approach allows for both quantitative interaction modifications and topology changes . To investigate this difference further , we used our model to evaluate differences in the contribution of weight modifications vs topology changes . We found that increasing the relative importance of topology changes ( by increasing the parameters for addition , ρ , and deletion , ϕ ) did not qualitatively change our results and in particular , did not create persistent sensitive nodes in the network ( S14A and S14B Fig ) . A reduction in the relative use of topology changes ( by reducing ρ and ϕ ) also did not change results qualitatively ( S14C and S14D Fig ) . In conclusion , these analyses suggest that MVGs explain the major difference in outcomes , namely persistent evolvability nodes in the Kashtan and Alon model compared to distributed and labile evolvability nodes in our host-parasite coevolution model .
Robustness—defined as tolerance to perturbations such as mutations and environmental fluctuations—is a pervasive feature of biological systems [17 , 38] . Previous models of gene regulatory networks have shown that mutational robustness evolves under conditions of stabilizing selection [34 , 38] . However , under more realistic scenarios , such as coevolution , evolvability may be advantageous . It is unclear though , how sensitivity and robustness will evolve and in particular how they will become distributed throughout a regulatory network . To investigate this we developed a two-population ( host-parasite ) model of antagonistic coevolution . Although previous studies [22–24] had investigated the evolution of sensitivity in networks under fluctuating environmental conditions , a key novelty of our model is that the fitness landscapes are emergent properties of the inter-population interactions . This approach avoids the need to impose a changing environmental regime externally . Furthermore , the pace of evolution is dictated largely by the model’s ability to adapt . Self-contained models such as these represent a step towards open-ended evolutionary models that will be critical in the longer term to understanding how biological complexity evolves . We found that sensitivity increases after the initiation of coevolution and becomes highly distributed throughout the network . At the same time , the remaining ( non-sensitive ) parts of the network evolve to become robust . Interestingly , genetic diversity evolves to be higher under antagonistic coevolution than under stabilizing selection . There are two obvious sources of diversity in this case . Firstly , in the non-sensitive parts of the network , robustness facilitates the accumulation of genetic variation via a well-understood mechanism [41] . Secondly , because sensitivity is distributed across the network , there are many different ways in which mutations cause phenotype inversions , contributing to diversity particularly after several rounds of selection . If sensitivity were not distributed , but were focused on a particular “evolvability hotspot” , genetic diversity could in principle be far lower . We found that robustness evolves in the parts of the network that are not involved in phenotype inversion . Interestingly , this robustness evolves more easily under asexual reproduction ( Fig 2B ) than sexual reproduction ( S6B Fig ) . Generally speaking we found that under sexual reproduction , a combination of higher network density , stronger selection pressure and/or larger population size was required in order to attain levels of robustness comparable to the asexual case , suggesting that recombination load is having nontrivial effects under sexual reproduction . In support of this , theoretical population genetic studies investigating the evolution of recombination [42 , 43] have shown that asexual reproduction will be favored over sexual reproduction under antagonistic coevolution when the two modes are allowed to compete . At the same time , a previous study using a similar network model to ours [44] , but having a single population under conditions of stabilizing selection , demonstrated that recombination load evolves to minimal levels under stabilizing selection . Thus it would appear that recombination load evolves to be higher under antagonistic coevolution than under stabilizing selection . We found that coevolutionary selection drives networks to evolve labile sensitivity such that evolvability and robustness are continuously redistributed throughout the network . Sensitive points within the network cause a phenotype inversion when mutated ( from S^→1−S^ ) , but the mutation by definition also changes the genotype , in particular by causing a change wij→wij' . Assuming the lineage continues through another phenotype inversion ( from 1−S^→S^ ) and wij' is not mutated again , then wij' will most likely no longer be sensitive . However , each time a sensitive point in the network is “used up” , a new sensitive point emerges elsewhere , thus maintaining overall sensitivity at approximately constant levels . We refer to this process as whack-a-mole sensitivity , named after a fun park game in which targets are removed from one place only to reappear elsewhere . A comparable whack-a-mole process also appears to occur with meiotic recombination hotspots in mammals [45 , 46] . During meiosis , recombination breakpoints are frequently initiated at DNA motif hotspots recognized by the PRDM9 protein . However , DNA repair mechanisms cause hotspots to be preferentially lost in the gametes of heterozygote ( hotspot/non-hotspot ) individuals and the net effect is for recombination hotspots to be lost over time . However , by means that are still not well understood , the overall number of hotspots ( in humans for example ) remains approximately constant while the positions of recombination hotspots are transient and vary within humans [47 , 48] , suggesting there must be a mechanism for generating new hotspots to replace those that have been lost , i . e . , a whack-a-mole process . Broad distributions of mutations have been observed in antibiotic resistance , for example in bacteria which produce extended-spectrum beta-lactamase ( ESBL ) enzymes [49 , 50] . In this case , many distinct point mutations occurring in ESBL genes such as TEM–1 and SHV–1 transform the active site of the enzyme . More than 330 ESBL variants including TEM- and SHV- type variants have been reported [50] . Whack-a-mole sensitivity may explain these rapidly expanding mutations in genes encoding ESBLs , thus helping to predict the evolution of resistance . In our model the ongoing phenotypic inversions are dependent on successive mutations that accumulate across many different loci . Because we found that sensitive nodes are labile ( whack-a-mole sensitivity ) , this means that over time similar mutations at a particular gene regulatory interaction might have distinct phenotypic effects . For example , a phenotype inversion ( S→1−S ) might be caused by a mutation at a particular sensitive site wij . The original phenotype S might then be restored by a reverse mutation at the same site . However , because sensitivity is distributed across the network , the restored phenotype S is more likely to arise through a mutation at some other site different than wij . Indeed several generations may pass before this reverse mutation occurs and by that time other mutations may have accumulated in the network . In this new genetic background the “reverse” mutation may no longer have the same effect . This is a clear example of serial epistasis—the dependency of mutational effect on the genetic background established by previous mutations [51] . A widely-cited study of serial epistasis in a natural population involves the evolution of resistance to the insecticide diazinon in populations of Australian sheep blowfly [52 , 53] . Here , an early resistance mutation arose conferring higher fitness in the presence of insecticide , but lower fitness compared to wildtype in the absence of insecticide . A second mutation then evolved to ameliorate the deleterious mutation , thus restoring fitness to wildtype levels for the double mutants . A key issue in evolutionary biology is understanding the extent to which epistasis , and in particular serial epistasis , determines the path of evolutionary change [51] . Such evolutionary constraints have been shown clearly at the level of individual proteins , for example , in a classic study of the evolution of novel function in vertebrate steroid receptors [54] , the authors evaluated experimentally the inferred ancestral proteins leading to the separate evolution of mineralocorticoid and glucocorticoid steroid receptors . They found that structural interactions imposed constraints that determined a specific ordering for the observed evolutionary substitutions . At the same time , the importance of serial epistasis in larger-scale systems such as regulatory networks is less well understood [55] . Our results suggest that whack-a-mole sensitivity will evolve as an emergent property of the network when there is distributed sensitivity and the serial epistasis effects that come with it . Taken together , under conditions of strong antagonistic coevolution , sensitivity in gene regulatory networks evolves to be broadly distributed and highly labile . Our results suggest there will be no central network elements that determine phenotype changes in the long term . Previous studies had found that network modularity could evolve in the context of alternating environments comparable to those that emerge in our model [22 , 23] . A modular network architecture can facilitate phenotype switching by perturbing key interaction ( s ) between modules [26] . However , we observe an entirely different mechanism based on sensitivity in which modularity does not play a role . When we compared with the Espinosa-Soto model [23] , we found that modularity did not evolve even when we adopted many model features , and perhaps the most relevant difference with that model lies in the nature of the perturbations ( S11 Fig ) . Modularity may be more likely to evolve in the face of environmental perturbations than in networks faced predominantly with mutational perturbations [25 , 26 , 56] . When we compared with the Kashtan model [22] , we found that introducing Modularly Varying Goals ( MVGs ) could , to some extent , drive the evolution of persistent sensitive interactions ( S12B Fig ) , although network modularity did not increase ( S12C Fig ) . As we have observed , distributed sensitivity offers the advantage of allowing a large number of mutations throughout the network to generate phenotype changes . If a network has many different regulatory interactions that enable rapid adaptation via point mutations , the network does not have to mutate a specific interaction back-and-forth in order to repeat the process .
The model largely follows previously published models [34–37] with the exception of selection , which here depends on interactions between the host and parasite populations . In our model , a genotype is represented as a matrix ( W ) where the elements wij describe the regulation of the i−th gene by the j−th gene product . Positive matrix entries represent activation , negative entries represent repression and zeros indicate no interaction . Gene expression is represented by a vector S ( t ) containing elements Si ( t ) representing the expression level , in the range ( 0 , 1 ) , over time t of the i−th gene . Initial gene expression , S ( 0 ) , is given as a random binary vector of 0 and 1 expression levels . Gene expression dynamics are determined by the difference equation S ( t+1 ) = σ ( W S ( t ) ) where σ ( x ) =11+e−ax is a sigmoid function . The steady state gene expression , S^ , is the phenotype and individuals not reaching steady state have zero fitness . The evolutionary simulation is initiated by creating a founder individual for each population in the form of a random matrix W of regulatory interactions containing non-zero elements with probability c , drawn from a Normal distribution , N ( 0 , 1 ) . The founder is copied to form the initial population of size M . Each population undergoes cycles of reproduction , mutation and selection . In the case of sexual reproduction candidate offspring are produced by inheriting a row in the matrix W at random from either parent . Here each row i represents regulatory interactions of the set of cis-regulatory elements controlling the expression of gene i . Row-wise inheritance implies inheritance of cis-regulatory regions and free recombination among loci . Under asexual reproduction , random parent genotypes are simply cloned . Following [36] , mutations apply to the genotype of each offspring , W and may cause addition of new network interactions ( when element wij = 0 ) or deletions and modifications ( when wij ≠ 0 ) . The mutation frequency per genotype is constant ( μ ) . Mutations lead to either addition ( ρ ) , deletion ( ϕ ) or modification ( δ ) of interactions . The addition and deletion rates are set to ensure that network density does not change from its initial value ( See Parameters section below ) . In the selection step , the interaction between host and parasite populations determines a distinct fitness function for each population , as described in the main text . Unless otherwise stated , the simulation results used the following parameter values: number of genes , N = 10; population size , M = 200; mutation rate per genotype , μ = 0 . 1; selection strength , α = 0 . 1; asexual reproduction; network density , c = 0 . 5 . As described previously , [36] the network density c , will be at steady state when its difference in time , Δc ( t ) = c ( t ) −c ( t−1 ) = μ ( α ( 1−c ( t ) ) −ϕc ( t ) ) /N2 is zero . We therefore chose the parameters for addition ( ρ = 0 . 025 ) and deletion ( ϕ = 0 . 025 ) that satisfy Δc ( t ) = 0 . Given these parameters , modifications are set to ( δ = 1−ϕ ) . As described above , a mutation is defined as the replacement of one element wij ( i , j = 1 , … , N ) with a random number drawn from a Gaussian distribution , N ( 0 , 1 ) if the interaction is either modified or added and with zero if the interaction is deleted . The sensitivity score is calculated by estimating the expectation of a phenotype inversion given a random mutation . This involves evaluating whether a mutation that would change wij→l will generate a phenotype inversion ( k ( l ) = 1 ) or not ( k ( l ) = 0 ) . Because the probability of the mutation wij→l follows a continuous Gaussian distribution f ( l ) , we employ a discrete approximation given by evaluating f ( l ) at 2L/δ+1 positions across the range [−L , L] separated by small intervals of size δ . More formally , the sensitivity score of an interaction ( wij ) in a network is measured as SSij=∑l=−LLδ⋅f ( l ) ⋅k ( l ) where l∈{−L+n⋅δ | −L+n⋅δ≤L , n∈ℤ*} ( = the range of mutation: wij→l ) , f ( l ) = 1σ2πe−l22σ2 ( normal distribution probability density function with mean = 0 ) and k ( l ) = 1 if the phenotype is inverted by the perturbation wij→l , otherwise k ( l ) = 0 . We consider a phenotype as inverted if the L1 distance ( ‖X−Y‖1=∑i=1N| xi−yi | ) between the original phenotype and a perturbed phenotype by the wij→l mutation , excepting those Nb genes that have basal expression ( si = 0 . 5 ) due to not having inputs , is greater than pflip ( N−Nb ) . For all results reported we used pflip = 0 . 9 and δ = 0 . 02 , σ = 1 and L = 3 , which covers the range of 99 . 73% of possible mutations at wij . The SS of a genotype ( W ) is the average of SSij for all elements wij , i . e . , SS=∑i=1N∑j=1NSSijN2 . The overlap in sensitivity , which describes the similarity of two genotypes , u and v , is measured as the Jaccard index , J ( u , v ) = |Au∩Av||Au∪Av| , where Au is the set of sensitive interactions in u for which SSij>0 and Av is the set for genotype v . We calculate J ( u , v ) for an individual ( u ) and its ancestor ( v ) of the same phenotype . Comparing with an ancestor of the same phenotype is fairer than comparing with an inverted phenotype . Because it is not possible to guarantee that an ancestor at a particular previous generation will have the same phenotype , we chose the closest ancestor having the same phenotype within a window of size 10 ( i . e . , assuming intervals of size 100 , ancestors in ranges of 1–10 , 11–20 , . . , 91–100 generations previous ) . As a null model , sensitive interactions , with SSij>0 are randomly redistributed in the networks . The overlap in sensitivity for the null model is calculated in the same way and the mean overlap for 100 null models is used ( Fig 3A ) . Assuming that a network has x interactions with sensitivity score SSij>0 and the sum of these x sensitivity scores is H . For the null model we used a standard string cutting method that generates x numbers such that their sum equals H . This was implemented by choosing x−1 random numbers in the range ( 0 , H ) and then calculating the distances between the adjacent numbers including the end points 0 and H . These distances ( of which there are x ) , are random numbers which are distributed according to a Dirichlet distribution , and whose sum is H . As we did for the original network , we calculated the standard deviation ( SD ) of these x randomly distributed sensitivity scores . We repeated this process 100 times and compared the mean value with the SD of the original network ( S5A Fig ) . To quantify environmental robustness we perturbed initial gene expression 500 times for each individual in the population by changing si→1−si at a rate 0 . 01/gene ( S7A Fig ) and 0 . 2/gene ( S7B Fig ) . We then calculated the phenotype distance D ( S1 , S2 ) =∑i=1N|S1i−S2i|N between unperturbed ( S1 ) and perturbed ( S2 ) phenotypes excluding phenotype inversion cases as the measure of environmental robustness . The modularity measure we used is taken from [57] . We restate the definition as follows: Given a graph G ( V , E ) , where |V| = k , |E| = l , the vertices of G ( V , E ) can be clustered into n clusters , C = {C1 , C2 , … , Cn} , 1≤n≤k . Modularity is defined as Q ( C ) =∑i=1k[ | E ( Ci ) |l− ( ∑v∈Cideg ( v ) 2l ) 2 ] where E ( Ci ) = {{vh , vt}∈E|vh , vt∈Ci} and deg ( v ) = |{∀vt≠v|{v , vt}∈E}| . We adapted the ( Espinosa-Soto ) model described in [23] to our antagonistic coevolution model as follows . The previous study used two target phenotypes simultaneously , each of which has a conserved part and a distinct part as input gene expressions . To emulate this in our model , we assigned half the genes to be under stabilizing selection and the other half under coevolutionary selection . The set of genes under stabilizing selection has the same target phenotype throughout the simulation whereas the other is under coevolutionary selection as with the model described in the main text . To represent environmental perturbations , the input gene expression is perturbed by changing si→1−si at a rate 0 . 15/N as described in [23] , for 400 experiments . The fitness function of an individual is f = 1−e−3γ , γ=∑i=1400 ( 1−Di/Dmax ) 5/400 , where Di is the Hamming distance between the target phenotype and a new phenotype from i th perturbed input . For the model version that assigns half of the genes to be under stabilizing selection and the other half to be under coevolutionary selection but does not include environmental perturbations , we calculate two types of fitness for stabilizing and coevolutionary selection respectively: fs= 1−e−dα , d=∑i=1Ns ( Soptimal ( i ) −S^ ( i ) ) 2/ ( Ns⋅ζ ) and fc={e−1−dα , hoste−dα , parasite , d=∑i=1Nc ( Santagonist ( i ) −S^ ( i ) ) 2/ ( Nc⋅ζ ) . ζ = 1 for continuous expression levels and ζ = 4 for discrete ( -1 , +1 ) expression levels . Ns is the number of genes under stabilizing selection and Nc is the number of genes under coevolutionary selection . Survival requires both fs and fc to exceed a uniformly-distributed random value in the range [0 , 1] . | Robustness , defined as tolerance to perturbations such as mutations and environmental fluctuations , is pervasive in biology . Previous models of gene regulatory networks have shown that robustness can evolve when the phenotype is under evolutionary pressure to remain constant ( stabilizing selection ) . But in more realistic scenarios such as coevolution , it may be advantageous to evolve sensitivity , i . e . for some mutations to change the phenotype . We developed a two-population ( host-parasite ) model to investigate how robustness and sensitivity become distributed within a network under antagonistic coevolution . We found that sensitivity follows a pattern , similar to that of the game “whack-a-mole , ” in which sensitive sites mutate , thus becoming insensitive , but new sensitive sites emerge to take their place . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [] | 2015 | Antagonistic Coevolution Drives Whack-a-Mole Sensitivity in Gene Regulatory Networks |
Long intergenic non-coding RNAs ( lincRNAs ) are appearing as an important class of regulatory RNAs with a variety of biological functions . The aim of this study was to identify the lincRNA profile in the dengue vector Aedes aegypti and evaluate their potential role in host-pathogen interaction . The majority of previous RNA-Seq transcriptome studies in Ae . aegypti have focused on the expression pattern of annotated protein coding genes under different biological conditions . Here , we used 35 publically available RNA-Seq datasets with relatively high depth to screen the Ae . aegypti genome for lincRNA discovery . This led to the identification of 3 , 482 putative lincRNAs . These lincRNA genes displayed a slightly lower GC content and shorter transcript lengths compared to protein-encoding genes . Ae . aegypti lincRNAs also demonstrate low evolutionary sequence conservation even among closely related species such as Culex quinquefasciatus and Anopheles gambiae . We examined their expression in dengue virus serotype 2 ( DENV-2 ) and Wolbachia infected and non-infected adult mosquitoes and Aa20 cells . The results revealed that DENV-2 infection increased the abundance of a number of host lincRNAs , from which some suppress viral replication in mosquito cells . RNAi-mediated silencing of lincRNA_1317 led to enhancement in viral replication , which possibly indicates its potential involvement in the host anti-viral defense . A number of lincRNAs were also differentially expressed in Wolbachia-infected mosquitoes . The results will facilitate future studies to unravel the function of lncRNAs in insects and may prove to be beneficial in developing new ways to control vectors or inhibit replication of viruses in them .
Dengue and Zika viruses are related mosquito-borne viruses that have a common primary vector , Aedes aegypti and infect millions of people worldwide [1 , 2] . Recent outbreaks of Dengue and Zika in South America pose a serious risk for other tropical regions in the world as Ae . aegypti is one of the most abundant mosquito species in these areas [2] . Although certain vaccines have been licensed in some countries , there are no efficient specific therapeutics available for either diseases , hence , the best protection against their global spreading is an efficient vector control program [3 , 4] . The genome sequence of Ae . aegypti is available , however , it has not been fully annotated . Only 2% of its large genome ( 1 . 376 Mb ) has been annotated as protein coding genes and it reflects the presence of great proportions of non-coding transcripts as well as repetitive elements [5] . Transcriptomic changes , including those of non-coding transcripts , could provide a genome scale insight into host-pathogen interactions . Previous studies identified a series of small ncRNAs in Ae . aegypti and demonstrated their interaction with arboviruses [6–9] , but our knowledge about their long ncRNAs is limited . RNA transcripts longer than 200 nucleotides , which do not contain an open reading frame of longer than 100 amino acids , are simply defined as long ncRNA [10] . Generally , they are classified by their location relative to their neighboring protein-coding genes and include the long intergenic ncRNA ( lincRNA ) , intronic lncRNA , antisense lncRNA and enhancer RNA [10] . Although a number of mammalian lncRNAs have been characterized and identified in the last few years , genome-wide identification of this class of ncRNAs has only recently become possible with the arrival of deep sequencing technologies . An expanding body of evidence reveals that lncRNAs , once described as dark matter , are involved in many biological processes such as genomic imprinting and cell differentiation [11] . They also play important roles in epigenetic and non-epigenetic based gene regulation [12] . Relatively , little is known about their involvement in activation and differentiation of immune cells , but new discoveries have revealed the involvement of lncRNA in defense systems [13] . Previous works have also outlined their quick responses to different stimuli and stress factors [14–17] . In addition , it has been shown that some lncRNAs enhance virus replication or decrease antiviral immunity [18] . Although in most host-virus interaction studies typically protein-coding genes have been the center of attention , there are few examples of virus and host lncRNA interactions in human and mouse models [18 , 19] . For instance , Hepatitis B virus ( HBV ) infection altered lncRNA profiles in patients , with about 4% of human lncRNAs showing more than 2-fold changes in HBV infected liver tissue [20] . Winterling et al . ( 2014 ) identified a virus inducible lncRNA , which is induced by vesicular stomatitis virus and several strains of Influenza A virus ( IAV ) [18] . The sequence and structure of lncRNAs are important in their function , in particular for their interaction with DNA , RNA , or proteins . In case of extensive base-pairing of lncRNA with target mRNA , translation can be stabilized , while partial base-pairing may accelerate mRNA decay or inhibit translation of the target mRNA [21] . It has been shown that some lncRNAs interact with other small ncRNAs such as miRNAs . For example , in silkworm , 69 lncRNAs originating from 33 gene loci , may serve as miRNA precursors , and 104 lncRNAs may function as competing endogenous RNAs ( ceRNAs ) [22] . LncRNAs are also targeted by miRNAs similar to mRNAs and reduce their stability . They may also act as sponge or decoy of miRNAs , and compete with miRNAs for binding to mutual target mRNAs [21] . In insects , only a few genes have been experimentally annotated as lncRNA . It has been estimated that more than 5000 loci potentially encode non-coding transcripts in Drosophila melanogaster , however , just seven loci ( bxd , Hsrω , pgc , roX1 , rox2 , sphinx and yar ) have been annotated as functional regulatory lncRNAs by experimentally derived data [23 , 24] . We recently found that a number of lncRNAs in Plutella xylostella , a pest of cruciferous plants , were linked to the insect’s resistance to insecticides and might be involved in detoxification processes [14] . Jenkins et al ( 2015 ) identified 2 , 949 lncRNAs in the malaria mosquito vector , Anopheles gambiae , using RNA-Seq data [25] . They showed that in various Anopheles species , lncRNAs have considerably lower sequence conservation as compared with protein-coding genes . In another study , it has been shown that 43% of total midgut transcripts of An . gambiae are lncRNAs and 32% of them showed some level of homology to other species [26] . The current study generated a comprehensive list of Ae . aegypti lincRNAs , which will be a complement to the other ncRNAs ( microRNAs and piRNAs ) that have already been discovered in this medically important species . This work also helps to improve the present annotation of the genome of Ae . aegypti . We also examined the expression pattern of some selected lincRNAs in response to microbial challenge namely dengue virus serotype 2 ( DENV-2 ) and Wolbachia infections to identify potential immune-related lincRNAs in Ae . aegypti [27 , 28] . The results help better understanding of mosquito-pathogen interactions providing new insights on the potential role of lncRNAs as candidates for exploitation to inhibit replication of mosquito-borne viruses .
Previously sequenced RNA-Seq raw data of Ae . aegypti were downloaded from NCBI Sequences Read Archive and ArrayExpress Archive with accession numbers SRA048559 , SRA058076 , SRA244067 and ERP002530 [29–32] . Raw data were stripped of adapters using CLC Genomic Workbench version 7 . 5 . 1 and reads with quality score of above 0 . 05 and maximum 2 ambiguous sequences were retained for further analysis . The CLC Genomic workbench’s Transcript Discovery plugin was used for lincRNA discovery in the Ae . aegypti genome . New transcripts were identified by large gap mapping of 1 , 148 , 814 , 115 reads of 35 RNA-Seq libraries to the genomic reference ( AaegL3 . 3 ) . We implemented strict mapping criteria ( mismatch , insertion and deletion costs: 2: 3: 3 respectively ) . The minimum similarity and length fraction of 0 . 9 between a mapped segment and the reference were allowed as part of the mapping criteria . The large gap mapper algorithm also requires each mapped segment to include at least 10% of the read with minimum length of 17 bases . We considered a gap with maximum of 50 Kbp distance between mapped read segments to span the introns from RNA-Seq data . The annotations were generated by inspecting mapping of reads and identifying likely regions corresponding to genes , including their exons and splice sites . The algorithm scans each gap in the read mapping to explore whether the gap is assigned to a valid splice site or can be relocated to a valid splice site without cost . A rigorous filtering pipeline was developed to remove transcripts that may potentially encode proteins . The pipeline for Ae . aegypti lincRNA discovery is summarized in Fig 1 . We identified 75 , 069 potential genes using the CLC Genomic Workbench transcript discovery algorithm . The genes that were annotated as known Ae . aegypti protein-coding genes were discarded and 30 , 865 potential genes were also checked for any exon or intron overlap with other known Ae . aegypti protein-coding genes . We selected 22 , 079 sequences , which were located more than 1kb away from any other known transcripts , for finding putative open reading frames ( ORF ) . All possible six frames were produced for all selected sequences and then the translated sequences were subjected to a domain search to identify any putative conserved protein domains through Pfam v27 . 0 database [33] . We discarded 8 , 795 sequences with potential ORF above 100 aa or conserved protein domains . The remaining sequences were submitted to a coding potential assessment tool ( CPAT ) , which utilizes a logistic regression model made with four sequence features: ORF size , ORF coverage , Fickett TESTCODE statistics and hexamer usage bias [34] . We applied the coding probability threshold of 0 . 3 , which led to discarding 376 sequences as putative coding RNAs . We also implemented an expression threshold on our data to strengthen the identification pipeline . Sequences with more than 10 mappable reads in at least 17 out of 35 RNA-Seq libraries were considered as valid sequences and were kept for the next step . Any possible similarity with other known proteins was found by using BLASTx algorithm against nr and Swiss port database ( E-value cut off 10−5 ) . Finally , 3 , 842 potential lincRNAs were identified and used for further analysis . To identify Ae . aegypti putative lincRNAs that are regarded as small RNA associated lincRNAs , we used the Blast algorithm to search for Ae . aegypti precursor miRNA sequences in the predicted Ae . aegypti lincRNA dataset . We also used publicly available small RNA libraries from DENV-infected and non-infected samples ( SRP026241 ) in this analysis for further characterization of lincRNA_1317 . All known Ae . aegypti miRNA sequences were mapped to lincRNA_1317 for possible best fitting using RNAhybrid , which is a tool for finding the normalized minimum free energy ( mfe ) of RNA . We did not allow G:U pairing in the seed region ( nucleotide 2–8 ) and required miRNA-lincRNA duplexes to have a helix in this region . Maximum 5nt were allowed as unpaired nucleotides in either side of an internal loop . LncTar algorithm [35] was used to explore any potential interaction between lincRNA_1317 and DENV-2 genome ( accession no . NC_001474 . 2 ) by finding the normalized mfe joint structure of two RNA molecules based on base pairing . The Ae . aegypti genome was annotated with the final list of lincRNAs and used as reference for RNA-Seq analysis in CLC Genomic Workbench . To measure the lincRNA normalized expression value , RPKM ( Reads Per Kilobase per Million reads ) was assigned for each library [36] . To find the differential expression pattern of lincRNAs in response to DENV infection , data from DENV-2 ( Jam1409 ) infected midgut and carcass tissues at 4 days post-infection ( dpi ) were compared with their corresponding control groups [30] . Baggerley's test , a count based statistical analysis was done on the data . The samples were given weights depending on their total counts . Based on the test the weights are obtained by supposing a Beta distribution on the proportions in a group , and estimating these , along with the proportion of a binomial distribution . We selected 20 potential lincRNAs with more than 4-fold change for further analysis with RT-qPCR in DENV-2 ( New Guinea C strain ) infected Ae . aegypti cell line ( Aa20 ) and screened their expression profile in Wolbachia-infected mosquitoes . Ae . aegypti infected with the wMelPop-CLA strain of Wolbachia ( +Wol ) and without Wolbachia ( -Wol , tetracycline-cured line ) were stocks produced previously [37] . For the experiments in this work , 4-day-old female mosquitoes were used from which total RNA was extracted with 6–10 adult mosquitoes for each biological replicates . Detection and validation of the relative abundance of selected lincRNAs was carried through lincRNAs’ specific primers using SYBR Green chemistry in real time PCR machine . Briefly , total RNA was extracted from cells using Qiazol reagent according to the manufacturer’s instructions ( Qiagen ) . The TURBO DNA-free kit ( Ambion , USA ) was used to remove possible genomic DNA contamination in RNA samples . First strand cDNA was synthesized from 2 . 5 μg of RNA using a poly-dT primer and Superscript III reverse transcriptase ( Life Technologies ) . qPCR primers were designed using primer design tool of NCBI [38] . QuantiFast SYBR Green PCR Master Mix with ROX was used to quantify the relative expression of lincRNAs between different treatments . Three independent biological replicates were considered along with three technical replicates for each treatment . Reactions were performed in a Rotor-Gene thermal cycler ( Qiagen ) under the following conditions: 95°C for 5 min , and 40 cycles of 95°C for 10s and 60°C for 30s , followed by the melting curve ( 68°C to 95°C ) . Melting curves were analysed to examine the specificity of amplification . Relative expressions were calculated using the Rotor-Gene software and the mosquito RPS17 as reference gene for normalization . Unpaired t-test was used to identify statistically significant differences . To check the functional importance of the identified novel lincRNAs , dsRNAs were synthesized to knockdown selected lincRNAs ( 2329 , 1613 and 1317 ) to check their effect on DENV replication . Briefly , primers with added T7 promoter sequence ( S1 Table ) were used to generate 250–600 bp PCR products from selected lincRNAs . Megascript T7 kit ( Ambion ) was used according to the manufacturer’s instruction to generate respective dsRNAs . To induce efficient RNA silencing , Ae . aegypti Aa20 cells were double transfected with dsRNAs against selected lincRNAs . Aa20 cells were re-suspended and ~3×105 cells were added to each well of a 12-well plate . Cells were allowed to settle for ~1 h , medium was removed and replaced with a transfection mixture consisting of 0 . 5 ml medium ( 1:1 Schneider medium and Mitsuhashi–Maramorosch with 10% FBS ) , 8 μl Cellfectin ( Invitrogen ) , and 5 μg dsRNA either for selected lincRNAs or GFP as control . Cells were also treated with 3 μg dsRNA 72 h after the primary transfection to increase the silencing efficiency of selected lincRNAs . Six hours after the secondary transfection , cells were infected at 1 multiplicity of infection ( MOI ) with DENV-2 ( New Guinea C strain ) . All the treatments were collected three days post-infection . RNA extraction and cDNA synthesis were carried out as above . qPCR was performed to confirm the knockdown and the effect of particular lincRNA knockdown on the genomic RNA of DENV-2 . Each treatment was repeated three times . All data from three biological replicates were subjected to one-way ANOVA statistical analysis . Brown-Forsythe test was used to check the equality of group variances and Tukey's multiple comparisons test was also used to examine significant statistical differences among treatments .
In total , 3 , 482 putative lincRNAs in 1 , 114 Ae . aegypti genome scaffolds were identified ( S2 Table ) . The Ae . aegypti lincRNA genes displayed a slightly lower GC content ( mean: 40 . 1% ) in comparison to 47 . 8% in their protein-coding gene sequences ( Fig 2A ) . The lower GC content or AT enrichment is a typical characteristic of lincRNAs and our findings are congruent with predicted lincRNAs in other species [14 , 39 , 40] . The majority of Ae . aegypti predicted lincRNAs are smaller than 3000 bases and their length distribution is represented in Fig 2B . These mosquito lincRNA candidates are notably shorter in length than protein-coding genes , demonstrating another well-known characteristic of lincRNA transcripts ( Fig 2C ) [41 , 42] . The majority of Ae . aegypti genome scaffolds contain less than five lincRNA loci ( ~80% ) , however , 23 of scaffolds ( 2% ) were enriched with more than 10 lincRNAs ( Fig 2D ) . The detailed information of these scaffolds , which contain the highest number of lincRNAs are summarized in Table 1 . We examined all the identified lincRNAs to determine their association with Ae . aegypti miRNA precursors and piRNA clusters . We found that the pre-miRNA sequences of aae-miR-2940 and aae-miR-285 are located in lincRNAs 1431 and 3299 , respectively . We could not detect any other pre-miRNA sequences identified in Ae . aegypti in the lincRNAs . Also , lincRNA 1978 and 792 are originated from two previously reported piRNA clusters [43] located at supercontig 1 . 478 and 1 . 98 , respectively . LincRNAs demonstrate low evolutionary sequence conservation even among closely related species [10 , 14] . We used the BLAST algorithm bit score to identify the level of similarity among Ae . aegypti lincRNA sequences with other closely related insect genomes such as Aedes albopictus , Culex quinquefasciatus and Anopheles gambiae ( Fig 3A ) . As expected , most of the identified lincRNAs showed high level of similarity with Ae . albopictus genome sequence and probably are genus specific . The E-value cut off 10−50 was applied to our screening with the BLAST algorithm to identify the conserved sequences . Although the Ae . aegypti lincRNAs shared high level of sequence similarity with the genome of Ae . albopictus , only 62 and 7 lincRNAs had sequence similarity with Cx . quinquefasciatus and An . gambiae , respectively ( Fig 3B ) . They were mostly limited to a single short region with high conservation . Following the identification of Ae . aegypti lincRNAs , we analyzed their transcript levels in DENV-2 infected mosquitoes . To produce the lincRNA profile of infected and non-infected mosquitoes , we re-analyzed previously published RNA-Seq data from Ae . aegypti midgut and carcass samples at 4 dpi [30] ( Fig 4 ) . 248 and 203 lincRNAs with fold changes above four were identified in the RNA-Seq libraries of midgut and carcass , respectively ( S3 Table ) . The majority of differentially expressed lincRNAs were considerably overexpressed in both samples . The abundance of only 32% of Ae . aegypti lincRNA candidates decreased in response to DENV-2 infection in the mosquito carcass sample . Thirty lincRNAs were differentially expressed in both examined samples . The transcription levels of 72 lincRNAs increased after infection while their expression could not be detected in the non-infected midgut tissue sample . We selected 20 candidates of those differentially expressed lincRNAs from RNA-Seq analysis data for further investigation . The relative expression of lincRNA candidates were examined by reverse transcription quantitative polymerase chain reaction ( RT-qPCR ) upon DENV-2 infection in Aa20 mosquito cells . Only significantly overexpressed lincRNAs after DENV-2 infection are represented in Fig 5 . Although we used Aa20 cells for the lincRNA expression assays , the expression patterns of almost all the examined lincRNAs ( 5 out of 6 ) were consistent with the RNA-Seq data ( adult mosquito carcass sample ) . We used a poly-dT primer to produce cDNA , which also confirmed that all of those identified transcripts have poly-A tails and therefore are true transcripts . Based on these results , significant increase in the transcription levels of a selected number of Ae . aegypti lincRNAs suggests their possible involvement in host-pathogen interaction but further investigations are required to confirm their roles in antiviral/immune responses . We also examined the impact of an endosymbiotic bacterium , Wolbachia , on some selected Ae . aegypti lincRNAs , which showed significant changes in response to DENV-2 infection . This gram-negative bacterium is transmitted maternally and potentially infects more than 40% of all insect species , manipulating its hosts using different strategies [44 , 45] . A fascinating aspect of Wolbachia infection is limiting replication of vector-borne pathogens in mosquitoes [45 , 46] . However , the mechanism ( s ) behind virus blocking is largely unknown . Here , we found that the transcript levels of several lincRNA genes significantly increased in Wolbachia-infected Ae . aegypti mosquitoes ( Fig 6 ) , which may lead to differential regulation of cellular protein-coding genes . Our previous studies showed that Wolbachia could manipulate host small ncRNAs such as miRNAs and piRNAs [47] . An overall induction of small ncRNAs between 18 and 28 nucleotides was also observed in Ae . aegypti cell line infected with wMelPop-CLA strain of Wolbachia [48] . It was assumed that the upregulation of small ncRNAs in infected cells may result in an enhanced immune response and activated RNAi pathway . However , the role of these modifications in the host lincRNA gene expression profile , and potentially in anti-viral responses , is unknown and may lead to the discovery of lincRNAs that could be utilized for inhibition of virus replication in mosquitoes . A recent study on mouse bone marrow-derived macrophage ( BMDM ) model reported a significant upregulation in 72 lincRNAs after treatment with the synthetic bacterial lipoprotein Pam3CSK4 , which acts through Toll-like receptor [49] . In another study , differential expression of approximately 500 annotated mouse lncRNAs was reported during infection with severe acute respiratory syndrome coronavirus [50] . Recently , it has been shown that honeybee lincRNAs are also differentially expressed during infection with various viruses such as sacbrood virus ( SBV ) and deformed wing virus ( DWV ) , but the biological significance of these lincRNAs is completely unknown [51] . Although exploring the in vivo functions of immune-related lincRNAs is one exciting area for future studies , the differential expression of some lincRNAs could simply be byproducts of mRNA biogenesis or changes in global transcriptional profile due to microbial challenges [52 , 53] . Struhl ( 2007 ) believed that the transcriptional machinery is not perfect producing RNAs that serve no purpose or have no significant role in infection [54] . On the other hand , there are several examples which have shown that lincRNAs could be potentially important factors in host antimicrobial responses , and may represent a new class of signaling molecules involved in innate immunity or provide a new layer in gene regulation . For instance , two interferon ( INF ) induced lncRNAs , which were upregulated by influenza and vesicular stomatitis viruses , regulate the expression of the antiviral factor tetherin in human HuH7 cells [55] . To confirm the role of DENV-induced lincRNAs in viral replication , we used RNAi-mediated silencing of two selected lincRNAs ( lincRNA_1317 and 1613 ) using dsRNA in Aa20 cells followed by DENV-2 infection . Only RNAi-mediated silencing of lincRNA_1317 led to enhancement of DENV-2 replication ( Fig 7A ) . Silencing of the lincRNA was confirmed by RT-qPCR ( Fig 7B ) . Interestingly , expression of Ae . aegypti lincRNA_1317 increased substantially following the progression of infection ( Fig 7C ) suggesting that this lincRNA might be involved in antiviral response . This idea is consistent with the finding that lincRNA_1317 was also highly overexpressed ( 2 . 33 fold ) in Wolbachia-infected mosquitoes as compared with non-infected mosquitoes ( Fig 6 ) . While there are no reports on the involvement of lncRNAs in host-pathogen interactions in insects , time-dependent over-expression of host lincRNAs in response to viral infection has been observed in humans . A recent study showed more than 80% of host cell lncRNAs were upregulated upon an adenovirus infection of human primary lung fibroblast cells [56] . Zhang et al . ( 2013 ) reported alterations of expression of cellular lncRNAs in HIV-1-infected T cells . Among differentially expressed lncRNAs , NEAT1 expression notably increased in infected cells . When NEAT1 was silenced , virus production was enhanced by increasing the nucleus-to-cytoplasm export of HIV-1 transcripts containing Rev-dependent instability element [57] . A significant induction in this lncRNA expression in response to influenza virus and herpes simplex virus infection has also been shown [58] . To further investigate the potential role of lincRNA_1317 in mosquito-pathogen interaction , we determined its association with host endogenous small RNAs and its possible direct interaction with DENV . Although this lincRNA is not located in any of the known piRNA clusters , the majority of mappable small RNA reads to its sequence are in the range of 26–29 nt ( S1 Fig ) . However , there was no difference in the mapping pattern and mapped read length distribution when reads from DENV-infected and non-infected small RNA libraries were mapped to lincRNA_1317 ( S1 Fig ) . It has been shown that piRNA-like small RNAs have a large impact on lincRNA transcriptome [57] , but our knowledge about the function of piRNA-mediated lncRNAs is still limited . Recently , it has been reported that piRNAs derived from transposons and pseudogenes facilitate the degradation of lncRNAs in mouse late spermatocytes [57] . Next , we hypothesized that Ae . aegypti lincRNA_1317 response to microbial challenge could be due to cross-regulation between miRNAs and the lincRNA . Ae . aegypti miRNA recognition elements on lincRNA_1317 were identified by calculating the normalized minimum free energy ( mfe ) of hybridization for each Ae . aegypti miRNA and lincRNA_1317 using RNAhybrid core script . Binding site enrichment was detected for a few miRNAs with more than two recognition elements ( Table 2 ) . For instance , more than four recognition sites were predicted for miR-278-5p and miR-252-3p on lincRNA_1317 . We also identified some hot spots for miRNA recognition sites on lincRNA_1317 , which may allow multiple miRNAs to bind to the same regions ( S2 Fig ) . miRNAs can reduce lincRNA stability by targeting their transcripts similar to mRNAs . Also , lincRNAs with multiple recognition sites may actually be competitive inhibitors of miRNA function and stopping them from binding to their genuine targets by sequestering them [21] . Although the mfe for some of those miRNA-lincRNA recognition sites suggests high probability of a binding event , further experimental investigations are required to validate this interface . We also used LncTar algorithm to predict any direct interaction between lincRNA_1317 and DENV-2 genome . One potential interaction was predicted in the region 1–3370 of lincRNA_1317 and the region of 3210–6579 of DENV-2 genome with mfe of -61 . 73 ( normalized dG -0 . 0184 ) . This tool has accuracy rate of 80% [35] , but does not consider the tertiary structure of RNA , which could play a role in RNA–RNA interactions , and therefore further studies are required to validate any potential interaction . The involvement of lincRNA_1317 in host response to viral infection might be through its interactions with regulatory proteins that are involved in epigenetic changes by directly interacting with chromatin modifying enzymes or DNA binding proteins such as transcription factors . This interaction has been shown in several examples in mammalian systems , including host-virus interactions in which lncRNAs mediate antiviral responses by controlling the expression of immune-related genes ( reviewed in [58] ) . Although our knowledge of the biological function of this class of ncRNAs in mosquitoes is still limited , the results generated from this study will facilitate forthcoming explorations of lincRNA functions in insects . Clearly , further research is required to provide concrete experimental evidence to support the role of lincRNA_1317 or any other Ae . aegypti lincRNAs in host-pathogen interaction . With advances in technology , the mosquito lincRNA-protein interactions can be identified using high-throughput sequencing of immunoprecipitated RNA after cross-linking ( CLIP-Seq ) . Further , functional studies could be carried out to characterize immune-related lincRNAs . The involvement of lincRNAs in pathways associated with responses to viral infection and cellular stress makes them interesting candidates as potential targets for manipulation to inhibit virus replication or control vector populations . | Aedes aegypti is a major vector of several viruses such as dengue and Zika viruses . Understanding the intricate interaction of viruses with mosquito vectors and the factors involved in virus replication are essential for developing effective arbovirus control strategies . In this study , we report a comprehensive list of long intergenic non-coding RNAs encoded by the genome of Ae . aegypti for the first time . In addition , we show that a number of these long non-coding RNAs are differentially expressed in mosquitoes infected with dengue virus , which could be involved in DENV-mosquito interaction . The outcomes provide a new avenue to explore mosquito biology and mosquito-virus interactions that may lead to the discovery of molecules that could be beneficial for vector manipulation . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"and",
"Discussion"
] | [
"invertebrates",
"dengue",
"virus",
"medicine",
"and",
"health",
"sciences",
"pathology",
"and",
"laboratory",
"medicine",
"gene",
"regulation",
"pathogens",
"microbiology",
"animals",
"long",
"non-coding",
"rnas",
"invertebrate",
"genomics",
"viruses",
"micrornas",
"rna",
"viruses",
"insect",
"vectors",
"aedes",
"aegypti",
"medical",
"microbiology",
"epidemiology",
"gene",
"expression",
"microbial",
"pathogens",
"viral",
"replication",
"disease",
"vectors",
"insects",
"animal",
"genomics",
"arthropoda",
"biochemistry",
"rna",
"mosquitoes",
"nucleic",
"acids",
"flaviviruses",
"virology",
"viral",
"pathogens",
"genetics",
"biology",
"and",
"life",
"sciences",
"genomics",
"non-coding",
"rna",
"organisms"
] | 2016 | Identification of Aedes aegypti Long Intergenic Non-coding RNAs and Their Association with Wolbachia and Dengue Virus Infection |
Visceral leishmaniasis ( VL ) is a common complication in AIDS patients living in Leishmania-endemic areas . Although antiretroviral therapy has changed the clinical course of HIV infection and its associated illnesses , the prevention of VL relapses remains a challenge for the care of HIV and Leishmania co-infected patients . This work is a systematic review of previous studies that have described predictors of VL relapse in HIV-infected patients . We searched the electronic databases of MEDLINE , LILACS , and the Cochrane Central Register of Controlled Trials . Studies were selected if they included HIV-infected individuals with a VL diagnosis and patient follow-up after the leishmaniasis treatment with an analysis of the clearly defined outcome of prediction of relapse . Eighteen out 178 studies satisfied the specified inclusion criteria . Most patients were males between 30 and 40 years of age , and HIV transmission was primarily via intravenous drug use . Previous VL episodes were identified as risk factors for relapse in 3 studies . Two studies found that baseline CD4+ T cell count above 100 cells/mL was associated with a decreased relapse rate . The observation of an increase in CD4+ T cells at patient follow-up was associated with protection from relapse in 5 of 7 studies . Meta-analysis of all studies assessing secondary prophylaxis showed significant reduction of VL relapse rate following prophylaxis . None of the five observational studies evaluating the impact of highly active antiretroviral therapy use found a reduction in the risk of VL relapse upon patient follow-up . Some predictors of VL relapse could be identified: a ) the absence of an increase in CD4+ cells at follow-up; b ) lack of secondary prophylaxis; and c ) previous history of VL relapse . CD4+ counts below 100 cells/mL at the time of primary VL diagnosis may also be a predictive factor for VL relapse .
Visceral leishmaniasis ( VL ) and human immunodeficiency virus ( HIV ) co-infection has emerged as a serious disease pattern [1] , [2] . HIV infection increases the risk of developing VL by 100 to 2 , 320 times in endemic areas [3] , [4] and , on the other hand , VL promotes the clinical progression of HIV disease and the development of AIDS-defining conditions [5] . Both infections switch the predominantly cellular immunological response from Th1 to Th2 through complex cytokine mediated mechanisms leading to a synergistic detrimental effect on the cellular immune response [6] , [7] , [8] . Other important findings related to HIV-Leishmania co-infection is a reduction in therapeutic response and high rate of relapse , which is the clinical deterioration after clinical improvement , observed in 25–61% of patients [9] , [10] , [11] , [12] . Although the term recurrence has also been used as synonym for relapse , recurrence applies to the finding of a parasite repeatedly . It is important to emphasize that neither of these two terms distinguishes parasitological persistence from re-infection . The poor therapeutic outcome , the high rate of relapse , the poliparasitic nature of VL in HIV-infected persons , as well as the atypical manifestations of the disease and the impaired access to health-care resources make HIV-infected individuals prone to enlarge the number of human reservoirs [13] . This concern is of utmost importance in Asia , where HIV and Leishmania co-infections are increasingly being reported in countries that are also facing parasite resistance to antimonial drugs [14] . Recent changes in the epidemiological patterns of HIV and Leishmania infections are likely to lead to a greater degree of overlap and greater risk of co-infection and they justify increased alertness . From a global epidemiologic viewpoint , two parallel trends are alarming: the ruralization of the HIV pandemic and the urbanization and spread of VL [1] , [15] . World Health Organization ( WHO ) [16] reports that the public health impact of leishmaniasis worldwide has been grossly underestimated for many years because notification was compulsory in only 32 of the 88 countries where 350 million people were at risk . The reported global incidence of co-infection is likely underestimated either because VL has not been included in the list of AIDS related opportunistic infection in all endemic areas . Before the widespread use of antiretroviral therapy , such co-infection was common in Europe [5] . The co-infection is now becoming proportionately more prominent in areas with poor access to antiretrovirals , such as Africa . In areas where it is available , highly active antiretroviral therapy ( HAART ) has changed the course of the HIV/AIDS epidemic and the outcome of associated opportunistic infections . However , evidence of relapse rate reduction in patients using HAART is conflicting [17] . This work is a systematic review of studies describing the predictors of VL relapse in HIV-infected patients .
This review was conducted on all papers published before July , 31 , 2010 . To ensure scientific rigour , the Preferred Reporting of Systematic Reviews and Meta-Analysis ( PRISMA ) guidelines [18] were used for systematic data synthesis . Studies were identified by a Medline/PubMed search using a combination of terms that has been shown to maximize sensitivity [19] . The search terms used are shown in Figure 1 . The LILACS and Cochrane databases were used for literature review using a Boolean combination of search terms . Additional reports were located using a manual search of references from retrieved papers . Two independent reviewers ( GFC and MRS ) initially checked the lists of titles and abstracts identified by this search to determine whether an article contained relevant data . Studies were considered eligible if they were presented in an original article , examined HIV-infected individuals over 14 years of age with a VL diagnosis , included follow-up after the leishmaniasis treatment and clearly analyzed predictors of relapse . There were no restrictions on the publication language , date of publication , use of secondary prophylaxis , or duration of follow-up in the study . We excluded studies evaluating fewer than ten cases and studies evaluating mixed populations of HIV-infected and uninfected subjects unless separated results for HIV patients were identified . The selected articles were read in full to confirm eligibility . Data were extracted directly from the full-length articles into structured tables containing all of the descriptive variables and relevant outcomes . The following information was extracted: country and period of enrollment; sample size; clinical characteristics of the included patients; study design; the number of excluded patients if specified; statistical analyses utilized; duration of follow-up and number of subjects lost to follow-up; outcome of interest; prognostic variables assessed in each study and quality of the regression model [20] , [21] , [22] . When data were available tests required for completion of the tables were performed . To summarize the results regarding secondary prophylaxis , the software Comprehensive Meta-Analysis Version 2 . 2 . 048 was used .
Table S1 summarizes the characteristics of the 1017 patients encompassed by the 18 included studies . The year of study publication ranged from 1989 to 2008 . The design of 8 of the studies examined was prospective . Fourteen studies were reported in Spain , two in Italy , and one in Ethiopia and one in France . Eight studies had an enrollment period exclusively after 1996 , when HAART became available . Twelve papers stated the proportions of patients receiving HAART involving two nucleoside reverse transcriptase inhibitors and one or two protease inhibitors or non-nucleosides reverse transcriptase inhibitors at VL diagnosis or at relapse or both . A large proportion of the patients in these studies ( 87 . 7% ) were male and most were young adults; the median or mean ages reported varied from 27 to 37 years ( Table S2 ) . In the 14 studies in which patients' presumed transmission route was known , 72 . 3% ( 420/581 ) of the infections were likely due to intravenous drug use . The median CD4+ T lymphocyte count ranged from 11 to 82 cells/mL . Most patients had an AIDS-defining condition [48] at the time of VL diagnosis ( 332/572 , 58% of patients ) . In the majority of the studies , the diagnosis of VL was established by direct demonstration of amastigotes ( by cytological study of Wright stains ) or by the observation of promastigote growth in samples cultured in specific media . In one study [49] , the VL diagnosis was supported either by positive results from Leishmania-specific PCR ( polymerase chain reaction ) of peripheral blood or bone marrow samples . Three studies [47] , [50] , [51] also included patients diagnosed by serologic tests ( direct agglutination , indirect immunofluorescence or rK-39 dipsticks ) . The drug used in the treatment of the primary episode of VL was reported for 89% of the treated patients . Of this total , 73 . 4% of cases ( 733 patients ) were treated with pentavalent antimonial drugs , 12 . 4% with amphotericin B deoxycholate ( 124 patients ) , and 2 . 1% ( 21 patients ) received amphotericin in lipid formulations . A minority of patients ( 1 . 2% ) received pentamidine isethionate and three papers included patients treated with miltefosine [47] or unconventional regimens such as a combination of allopurinol with an azole compound [50] , [52] . A test of cure ( staining with Giemsa stain and parasite culture or PCR ) at the end of treatment was carried out in 8 of 18 studies . In most of these studies , this control was performed for patients whose clinical response was uncertain . Secondary prophylaxis for leishmaniasis was reported in eleven studies . Three studies explored the impact of mono or dual antiretroviral therapy at VL diagnosis [47] or during the follow-up [50] , [53] on relapse . Only one [47] of these studies demonstrated a reduction in relapse rate compared with patients who did not undergo retroviral therapy . Similarly , only one [49] of four studies [10] , [49] , [51] , [54] that followed patients on HAART at VL diagnosis reported a reduction in relapse rate . HAART use on follow-up has also been studied in relation to risk of relapse and none of the five [9] , [51] , [52] , [54] , [55] studies showed reduction on VL relapse rate . Two studies [52] , [54] that evaluated VL prophylaxis without specifying the drug used noted a significant reduction in relapse . In a report of ten cases , Bossolasco et al . [55] showed that the relapse rate in patients groups with and without prophylaxis were 60% and 100% , respectively , but this difference did not reach statistical significance . Three studies evaluated specific prophylactic regimens ( antimony compounds [46] , [50] and liposomal amphotericin [50] ) and demonstrated reduction on VL relapse . Although the confidence intervals did not reach statistical significance , another author [56] concluded that lipid-complexed amphotericin prophylaxis also reduced the relapse rate . Finally , Laguna et al . [57] showed a trend towards ( p = 0 , 08 ) a reduction in VL relapse rate following treatment with pentamidine prophylaxis . A meta-analysis of results from all studies evaluating the impact of secondary prophylaxis is shown in Figure 3 . This analysis could consistently demonstrate that secondary prophylaxis reduces VL relapse rate . CD4+ lymphocyte count at VL diagnosis and follow-up has been studied in relation to risk of relapse . Nine articles [10] , [11] , [12] , [46] , [50] , [51] , [52] , [55] , [58] compared CD4+ lymphocyte cell counts at VL diagnosis between relapsing and non-relapsing patients as a continuous variable . Neither of these studies showed significant differences between these two groups . On the other hand , two studies [47] , [49] that compared relapse rate between patients with CD4+ count at VL diagnosis as a dichotomic variable ( above and below than 100 cell/mL ) noted that the arms with higher CD4+ counts had lower relapse rate . Similarly , an increase in CD4+ lymphocyte count at follow-up was protective against VL relapse in 5 of 7 studies [10] , [11] , [49] , [55] , [58] . In another study [12] , univariate analyses of CD4+ counts at follow-up revealed a trend towards a reduction in relapse ( p = 0 . 09 ) . Other variables explored in relation to relapse are shown in Table S3 . Factors such as age , route of HIV transmission , history of intravenous drug use , HIV viral load at VL diagnosis , various clinical findings , specific anti-Leishmania treatments given , time from VL diagnosis to the introduction of protease inhibitor therapy , HAART compliance , the presence of an AIDS-defining disease before VL diagnosis and the presence of serum anti-Leishmania antibodies were not substantially different between relapsing and non-relapsing patients . Tuberculosis co-infection [47] , hepatitis C virus co-infection [49] and an incomplete course of VL treatment [52] were evaluated in multivariate analysis and showed a statistically significant association of these conditions with the occurrence of relapse . Previous VL episodes were identified as risk factors for relapse in 3 studies , two of which were multivariate analyses . The statistical quality and the presentation of methods and results in many studies were poor . In nine studies , the Kaplan-Meier method was used in a univariate survival analysis to analyzed VL relapse . Three prospective studies and two retrospective cohort studies employed Cox regressions for multivariate analysis of independent predictors . One study randomized patients to compare prophylaxis ( liposomal amphotericin versus no treatment ) and performed multivariate analysis to compare relapse rates by logistic regression , including some predictors as covariates . None of these six studies mentioned collinearity assessment ( i . e . , a high degree of correlation between 2 predictive variables ) or developed a risk score for relapse based on their multivariable results . Also , none of the multivariate analyses reported a goodness-of-fit test of their models . Other studies analyzed isolated relapse predictors by univariate association tests in series of prospective or retrospective cases or in intervention studies .
Although we have made an extensive review , our analysis includes studies with different definitions of cure and different lengths of follow-up . Cure is seldom defined parasitologically in these studies and the difference between treatment failure and relapse is arbitrary in some studies . It is possible that some episodes of relapse in the group of patients in which parasitological cure were not documented by bone marrow examination were treatment failures rather than relapses . Moreover , re-infection was not distinguished from relapse in any paper . There is a high degree of heterogeneity in the evaluated populations as shown by the wide range of reported mortality ( 6 . 5% to 83 . 8% ) , treatment failure ( 0 to 47 . 6% ) and relapse rates ( 20% to 70% ) . These studies included patients with different degrees of immunosuppression , and different treatment and prophylaxis regimens . Also , there are differences in the study designs , the types of statistical methods used and the prognostic variables included in analysis . These variations may have resulted in patient selection bias or low statistical power , thus hampering a meta-analysis of all studied predictors of relapse . In spite of these limitations , we believe that the meta-analysis results of secondary prophylaxis are consistent , considering the available evidence . In addition , the quality of published reports was heterogeneous and usually poor . Despite these limitations , this review may assist clinicians in making decisions and may also help in the design of future studies . The results of this systematic review suggest there are identifiable predictive factors of VL relapse , such as previous episodes of VL relapse and lack of recovery of CD4+ lymphocyte numbers after primary visceral leishmaniasis . HAART did not produce the anticipated decrease in the incidence of VL relapses and more data is needed in order to better assess the evolution of VL in the HAART era . In contrast , secondary prophylaxis was shown to be protective against relapse . CD4+ count below 100 cells/mL at the time of VL primary diagnosis is a potential predictor of relapse . Based on these observations , a high-risk population might be identified and such patients might then be eligible for secondary prophylaxis . Strong surveillance will certainly contribute to improved data quality for decision-makers in this complex scenario . Randomized trials to compare the efficacy of different drugs and their role either in treatment or in prophylaxis are required . | Visceral leishmaniasis ( VL ) is the most serious form of an insect-transmitted parasitic disease prevalent in 70 countries . The disease is caused by species of the L . donovani complex found in different geographical regions . These parasites have substantially different clinical , drug susceptibility and epidemiological characteristics . According to data from the World Health Organization , the areas where HIV-Leishmania co-infection is distributed are extensive . HIV infection increases the risk of developing VL , reduces the likelihood of a therapeutic response , and greatly increases the probability of relapse . A better understanding of the factors promoting relapses is essential; therefore we performed a systematic review of articles involving all articles assessing the predictors of VL relapse in HIV-infected individuals older than 14 years of age . Out of 178 relevant articles , 18 met the inclusion criteria and in total , data from 1017 patients were analyzed . We identified previous episodes of VL relapse , CD4+ lymphocyte count fewer than 100 cells/mL at VL diagnosis , and the absence of an increase in CD4+ counts at follow-up as major factors associated with VL relapse . Knowledge of relapse predictors can help to identify patients with different degrees of risk , facilitate and direct prophylaxis choices , and aid in patient counseling . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"medicine",
"infectious",
"diseases",
"hiv",
"leishmaniasis",
"hiv",
"clinical",
"manifestations",
"neglected",
"tropical",
"diseases",
"viral",
"diseases",
"hiv",
"opportunistic",
"infections"
] | 2011 | Predictors of Visceral Leishmaniasis Relapse in HIV-Infected Patients: A Systematic Review |
Recent improvements in next-generation sequencing of tumor samples and the ability to identify somatic mutations at low allelic fractions have opened the way for new approaches to model the evolution of individual cancers . The power and utility of these models is increased when tumor samples from multiple sites are sequenced . Temporal ordering of the samples may provide insight into the etiology of both primary and metastatic lesions and rationalizations for tumor recurrence and therapeutic failures . Additional insights may be provided by temporal ordering of evolving subclones—cellular subpopulations with unique mutational profiles . Current methods for subclone hierarchy inference tightly couple the problem of temporal ordering with that of estimating the fraction of cancer cells harboring each mutation . We present a new framework that includes a rigorous statistical hypothesis test and a collection of tools that make it possible to decouple these problems , which we believe will enable substantial progress in the field of subclone hierarchy inference . The methods presented here can be flexibly combined with methods developed by others addressing either of these problems . We provide tools to interpret hypothesis test results , which inform phylogenetic tree construction , and we introduce the first genetic algorithm designed for this purpose . The utility of our framework is systematically demonstrated in simulations . For most tested combinations of tumor purity , sequencing coverage , and tree complexity , good power ( ≥ 0 . 8 ) can be achieved and Type 1 error is well controlled when at least three tumor samples are available from a patient . Using data from three published multi-region tumor sequencing studies of ( murine ) small cell lung cancer , acute myeloid leukemia , and chronic lymphocytic leukemia , in which the authors reconstructed subclonal phylogenetic trees by manual expert curation , we show how different configurations of our tools can identify either a single tree in agreement with the authors , or a small set of trees , which include the authors’ preferred tree . Our results have implications for improved modeling of tumor evolution and the importance of multi-region tumor sequencing .
The clonal evolution hypothesis in cancer states that cancer genomes are shaped by numerous rounds of cellular diversification , selection and clonal expansion [1 , 2] . Recent methods to characterize tumor clonal evolution can be divided into two broad classes—sample tree reconstruction and subclone tree reconstruction . The first class of methods models the history of clonal evolution in an individual as a phylogenetic tree with leaves being the individual’s tumor samples , yielding a relative temporal ordering and estimate of divergence between the samples [3–5] . The second class aims at reconstructing the history of clonal evolution as a tree , which summarizes lineage relationships between cellular subpopulations [6–9] . Until single-cell sequencing data is widely available , accurate high resolution modeling of tumor evolution [10] will likely remain exceedingly difficult , if not impossible . On the other hand , the coverage depth of current next generation sequencing experiments limits the number of cellular subpopulations or subclones detectable in tumor samples to a few ( approximately 5–10 ) that have undergone signficant clonal expansions [4 , 11–14] . Each of these subclones emerge from a parental population of cells by acquiring additional somatic mutations , and cells within each subclone can be assumed to be homogeneous . Modeling of subclone evolution often involves estimating the fraction of cancer cells harboring each somatic mutation i . e . , somatic mutation cellularity , which can be inferred from next generation sequencing read count data . For example , PyClone [15] employs a Markov Chain Monte Carlo method to identify groups of mutations with similar cellularities , and SciClone [16] uses variational Bayes mixture models to cluster somatic mutations by their read count frequencies , which can be a proxy for cellularities . Most recently , methods that couple the problems of somatic mutation clustering and phylogenetic reconstruction have emerged . PhyloSub applies a tree-structured stick breaking process that introduces tree-compatible cellularity values for mutation clusters [6] . A combinatoric approach based on an approximation algorithm for binary tree paritions [7] and a mixture deconvolution algorithm [8] have also been developed . However to our knowledge , most recently published studies of multi-region tumor sequencing continue to employ manual curation to construct a subclone phylogeny , after mutation cellularity has been estimated computationally [12–14] . We propose that progress in methods to reconstruct subclonal phylogenies will be substantially enabled by decoupling the problems of temporal ordering of subclones from that of mutation cellularity estimation . The SubClonal Hierarchy Inference from Somatic Mutations ( SCHISM ) framework described here can incorporate a variety of methods to estimate the cellularity of individual mutations , the cellularity of mutation clusters , and to build phylogenetic trees . First , we derive a novel mathematical formulation of assumptions about lineage precedence and lineage divergence in tumor evolution that have been fundamental to other subclone tree reconstruction methods [6–9] . Lineage precedence is modeled in terms of a statistical hypothesis test , based on a generalized likelihood ratio . Hypothesis test results are combined with lineage divergence assumptions and formulated as a fitness function that can be used to rank tree topologies , generated by a phylogenetic algorithm . In this work , we designed an implementation of genetic algorithms to build phylogenetic trees . However , the fitness function can also be combined with other approaches to phylogenetic tree reconstruction . The hypothesis test can be combined with any method to estimate mutation or cluster cellularities to infer their temporal orderings . We use simulations to evaluate the power of the hypothesis test and show that for many combinations of tumor purity , sequencing coverage , and phylogenetic tree complexity , the hypothesis test has good power ( ≥ 0 . 8 ) and Type 1 error is well controlled , when at least three samples from a patient are available . The simulations also confirm that the problem of subclonal phylogenetic tree reconstruction is underdetermined in many settings when the tumor sample count per individual is smaller than the number of subclones i . e , nodes in the phylogenetic tree . In these cases , we may see that the genetic algorithm identifies multiple equally plausible phylogenetic trees . However , when the problem is sufficiently determined , in general when the number of samples equals or exceeds the tumor sample count , the genetic algorithm reliably reconstructs the true tree , in most combinations listed above . Using data from three published multi-region tumor sequencing studies of murine small cell lung cancer [13] , acute myeloid leukemia [12] and chronic lymophocytic leukemia [14] , we show how SCHISM can be configured with a variety of inputs . For all samples in these three studies , SCHISM identified either a single tree in agreement with the tree reconstructed manually by the authors , or a small set of trees , which include the authors’ published tree .
Recent studies of small cell lung cancer ( SCLC ) in mice [13] , acute myeloid leukemia ( AML ) [12] , and chronic lymphocytic leukemia ( CLL ) [14] attempted to infer the subclonal phylogeny underlying tumor progression , based on sequencing of multiple tumor samples . All of the studies applied computational methods to cluster somatic mutations . In some cases mutation cluster cellularities were provided , while in others read counts or cluster mean variant allele fraction was provided . The authors did not use computational methods to reconstruct the subclonal phylogenies . We applied SCHISM to these datasets , using a variety of configurations . In cases where mutation cluster cellularities were available [13] , we used the hypothesis test on pairs of mutation clusters . If mean variant allele fraction for clusters was available [12] , we inferred cellularity as described in ( S1 Text:Eq . S4 ) and used the hypothesis test on pairs of mutation clusters . When only read counts and mutation cluster assignments were available [12 , 14] , we used our own naive estimator to derive cellularity values , and applied the hypothesis test to pairs of mutations . For all configurations , we constructed a precedence order violation matrix for all pairs of mutations ( POV matrix ) or all pairs of mutation clusters ( CPOV matrix ) and ran the genetic algorithm . This approach consistently identified phylogenies identical to those manually constructed by the authors as either the single maximum fitness tree or among a small set trees tied for the maximum fitness , in underdetermined cases . In the AML study , the authors did not construct phylogenies for three out of eight patients , but they predicted which of two general clonal evolution models best explained relapse in these three patients . For two of these patients , our phylogenies were in agreement with the authors’ clonal evolution model . For the third patient , our phylogenies suggested that either of the two clonal evolution models might explain relapse . Methods details are all described in Methods ( sections on Hypothesis test , Precedence Order Violation Matrix , Application to mutation clusters , Vote Aggregation and Subclonse size estimation ) and S1 Text ( sections on Naive mutation cellularity estimate and Cluster cellularity estimation ) .
Representing tumor evolution as a phylogenetic tree of cell subpopulations can inform critical questions regarding the temporal order of mutations driving tumor progression and the mechanisms of recurrence and metastasis . As the cost of next generation sequencing with high coverage depth decreases , many labs are employing multi-region tumor sequencing strategies to study tumor evolution . However , going from multi-region sequencing data to a subclonal phylogeny is a computationally challenging task and methods are still in their early days . Here we derived a novel framework to approach the problem . We described a statistical hypothesis test and mathematical representation of constraints on subclone phylogenies , based on rules of lineage precedence and divergence that have informed previous works in the field . We designed a new fitness function that can be used to constrain the process of subclone tree reconstruction . These tools comprise a flexible framework called SCHISM , which can be integrated with many existing methods for mutation cellularity estimation and phylogenetic reconstruction . Combined with a new implementation of genetic algorithms , we demonstrated the utility of SCHISM with simulations and by application to published multi-region sequencing studies . We were able to reconstruct the subclonal phylogenies derived by manual curation in these studies with high fidelity . Today’s multi-region sequencing studies may often have a limited number of tumor samples , due to restrictions on the number of biopsies likely to be performed for living patients . Our results suggest that even when only a few samples are available , more accurate estimates of mutation cellularity at higher purity and coverage increase the power of the SCHISM hypothesis test . A more subtle result is that the power and Type 1 error of the test also depend on the accuracy of the standard error estimates for cellularity values . The dependency can be seen directly in the derivation of the test statistic itself ( Methods:Eq 13 ) and indirectly in the ability of SCHISM to reconstruct complex subclone phylogenies in murine models of SCLC [13] . Although only two or three samples were sequenced from each mouse , the authors provided robust statistical estimates of mutation cluster cellularity and standard deviations . To our knowledge , our study is the first to apply genetic algorithms to the problem of subclone tree reconstruction . Current sequencing technologies limit discovery to approximately 5–10 major subclones in patient tumor samples , equivalent to phylogenetic trees with 5–10 nodes . It is likely that in the near future , improved technology will enable discovery of a larger number of subclones . The number of topologies for a tree with n nodes is equal to nn−2 [18] . Therefore it becomes increasingly difficult to use exhaustive enumeration over all topologies when n = 9 ( approximately 4 . 8 million topologies ) , n = 10 ( approximately 100 million topologies ) , and certainly for n > 10 . The genetic algorithm presented here enables heuristic searches over very large numbers of topologies and consequent evaluation of candidate phylogenetic trees , according to the extent to which they violated the rules of lineage precedence and divergence . However , the genetic algorithm will not always succeed when n is very large , and its success depends on the topology of the true tree and the distribution of mutation cluster cellularities ( S1 Text ) . Alternative heuristic approaches might also prove useful in this setting such as tabu search [19 , 20] , simulating annealing [21 , 22] , or iterated local search [23] . Currently , the topology cost component of the fitness function is informed by a statistical hypothesis test that addresses a dichotomous question about the ancestral relationship of two nodes . The mass cost is a numeric measure that quantifies the extent to which the lineage divergence rule is violated , and tree fitness depends on this numeric value , rather than on acceptance or rejection of a null hypothesis . We took into consideration the power loss that would result from a statistical test based on mass cost . Such a test would depend on the estimated cellularities of a parental cluster and the sum of cellularities of its child clusters ( Methods:Eq 31 ) . A possible null hypothesis could be that the difference between parental and sum of child cluster cellularities is non-negative ( no violation of lineage divergence rule ) , and where the absolute value of the difference measures the magnitude of the violation . This test would be underpowered compared to the topology cost test , because the expected confidence interval for the sum of mutation cluster cellularity values ( Eq 2 ) will be larger than that for a single mutation cluster . Thus , we chose to focus our hypothesis testing framework on the topology cost . The value of our current strategy to include a numeric measure of mass cost in the fitness function is explored in ( S1 Text and S1 Fig ) . The fitness function used in this work could be further improved by incorporating measures of mutation or mutation cluster importance , using knowledge about ordering of specific driver mutations based on tumor biology , synthetic lethality , or results from single-cell sequencing . The genetic algorithm used in this work could itself be improved by the addition of online termination criteria and adaptive modulation of its key parameters , such as crossover and mutation probabilities . A number of excellent methods to reconstruct subclonal phylogenies have been recently published [6–9] , and we believe all of them are likely to be useful to the cancer research community . A feature matrix comparing attributes of SCHISM to four other published methods is in S1 Table . Key contributions of SCHISM are that its genetic algorithm can handle more complex tree topologies than are tractable by brute force while retaining lack of in-built bias towards linear or branched tree topologies , its modularity , and its capability to integrate information from multiple samples ( biopsies from a patient ) in a new statistical framework . Finally , it is clear that under many circumstances , particularly when sample count is low and tree complexity is high , the problem of subclone tree reconstruction is underdetermined . It is likely that for at least some tumor types , the true subclone trees may be very complex . In the future , sequencing studies with a large number of samples per patient will be essential to accurately characterize these trees .
The hypothesis test and genetic algorithm used in this work are components in a general framework that decomposes the problems of mutation cellularity estimation , mutation clustering and subclone tree reconstruction ( Fig 6 ) . Given aligned reads from whole-genome , whole-exome or targeted deep next generation sequencing , any method for mutation cellularity estimation and/or clustering can be combined with the hypothesis test described in this section . If cellularities are estimated for a cluster of mutations , the test can be applied directly to temporal ordering of clusters . If cellularities are estimated for specific mutations , the test can be applied to infer temporal ordering of mutation pairs . Given assignments of mutations to clusters , a voting aggregation scheme can be used to order the clusters themselves . The precedence order violation matrix and cluster precedence order violation matrix summarize the output of the hypothesis tests . They can be used to visualize the statistical support for potential temporal orderings ( as in Figs 3 , 4 and 5 ) . Finally , a fitness function that depends on constraints for possible values of cluster cellularities ( mass cost ) and the results of the hypothesis test ( topology cost ) can be used to rank possible topologies of subclone phylogenetic trees . The fitness function is independent of the genetic algorithm search strategy proposed in this work . According to the infinite sites assumption [6–8] , each somatic mutation in a tumor arises only once throughout the history of the disease , and once a mutation occurs in a cell it is inherited by all descendants of the cell . It follows that given multiple tumor samples from an individual , a mutation may be present in varying proportions of the tumor cells in each sample , referred to as varying cellularity of the mutation across samples . SCHISM constructs a rooted phylogenetic tree to represent the history of tumor clonal evolution in an individual . Each tree node represents cells harboring a unique compartment of mutations , defining a subclone . Each edge represents a set of mutations , acquired by the cells in the child node and differentiating them from the cells in its parental node . The somatic mutations of each tumor cell then uniquely map it to one of the nodes in the tree . From the infinite sites assumption , we can infer that each mutation is uniquely assigned to an edge and the cells represented by a node harbor all mutations present in their parental node . Furthermore , a mutation present at a node cannot have cellularity greater than the mutations at its parental node , defining a lineage precedence rule . Also , the sum of mutation cellularities occuring in child nodes cannot exceed the mutation cellularity of their parent , because these mutations occur in mutually exclusive cellular populations , defining a lineage divergence rule . Many methods have been proposed to estimate mutation cellularities , and our framework can be used with any of these methods . In our reconstruction of subclone phylogenies from three multi-sample sequencing studies ( Results ) , cellularities were derived using ABSOLUTE and our own naive estimator ( S1 Text ) . Other methods such as PyClone or SciClone could also be applied . The Cluster Precedence Order Violation ( CPOV ) matrix can be generated by the following vote aggregation approach . Let the set of mutations assigned to cluster I be M ( I ) . Rows and columns of the POV matrix can be reordered so that mutations belonging to the same cluster are adjacent . Then the ordered interaction of any pair of clusters ( I , J ) is represented by a block of matrix entries with addresses P O V [ i , j ] ∀ i ∈ M ( I ) , ∀ j ∈ M ( J ) ( 25 ) The support for potential lineage precedence of cluster I to cluster J can be summarized by a vote of the matrix elements within the block , represented as an element ( I , J ) in a cluster-level POV matrix CPOV C P O V [ I , J ] = ∑ i ∈ M ( I ) , j ∈ M ( J ) P O V [ i , j ] | M ( I ) | · | M ( J ) | ( 26 ) where ∣M ( X ) ∣ denotes the number of mutations in cluster X . A genetic algorithm ( GA ) is a heuristic search inspired by the process of natural selection . In an initial generation , a set of random objects is created and their fitness with respect to a fitness criteria is evaluated . Next , objects from the initial generation are selected according to their fitness to be parents of the following generation , with a preference for the fittest parents . The parental objects reproduce themselves , and their progeny may harbor new variation . The process is repeated for either a fixed number of generations or until a pre-defined convergence criteria is reached . In our implementation , the GA searches through a space of phylogenetic tree topologies , ranking them with a fitness function that we derived based on our model assumptions . In the initial generation , we generate G0 = 1000 random tree topologies and evaluate the fitness of each tree . A sample of size 0 . 8 * G0 trees are selected for reproduction by a fitness proportional selection method [25] and their progeny are generated , using crossover and mutation operations ( Figs 7 and 8 ) . To increase diversity and avoid too fast convergence to a local optimum , 0 . 2*G0 random tree topologies are also generated . The following generation then consists of a mixture of the progeny of the previous generation ( s ) and new random trees , and the total number of trees is the same as in the previous generation , so that G1 = G0 . The process is repeated for a fixed number γ = 20 generations . For each generation , the trees selected to be parents are not limited to the previous generation only , but can be selected from any preceding generations . To avoid getting trapped in local optima , four independent runs of the GA are performed , each with 20 generations ( 1000 trees per generation ) , and the entire ensemble of trees sampled in the four runs is ranked by tree fitness . In this work , the number of mutation clusters and the cellularity of each mutation cluster in each sample is assumed to be known , and the GA is applied to explore the space of tree topologies with a given node count , including both linear and branched topologies . The simulations were designed to generate data compatible with a set of likely tree topologies and assess how well SCHISM could recover these topologies from the data . Given a tree topology , a simulation produces a set of tumor samples consistent with lineage relationships summarized in the tree . We assume that while the samples share these lineage relationships , each represents an independent instantiation of cellularity distributions over the edges of the tree . The variability among these simulated samples captures the stochastic process of preferential sampling of tumor cells in an individual’s multiple tumor samples . In each simulated sample , we model variant and reference read counts for mutations belonging to each edge in the tree , taking into account sequencing coverage depth , sample purity level and mutation cluster cellularity . Based on our modeling assumptions , it is straightforward to conclude that in each sample s , the fraction of tumor cells belonging to the subclone described by node n in a tree can be calculated as C p ( n ) s - ∑ q ∈ D ( n ) C q s ( 33 ) where C p ( n ) s is the cellularity of the mutation cluster associated with the upstream edge incident to node n , i . e . , p ( n ) in sample s , and ∑ q ∈ D ( n ) C q s is the sum of cellularities of mutation clusters associated with its set of immediate descendant edges D ( n ) . Each element of the precedence order violation matrix POV[I , J] is a binary indicator of whether the null hypothesis that mutation i can be ancestral to mutation j is rejected . Each element in the POV matrix is compared with its true value , given the correct tree topology . Performance is summarized by power and Type 1 error . The ability of the genetic algorithm to identify the correctly reconstructed subclone tree is assessed across multiple settings of key variables: tumor purity ( 0 . 5 , 0 . 9 ) , sequencing coverage depth ( 150X , 1000X ) , tree node count ( 3–8 ) , and tumor sample count ( 1–10 ) . For each node count , multiple tree instances ( alternate topologies for a given node count ) are considered ( S3 Fig ) . Then for each combination of settings , ten different replicates are run ( S4 Fig ) . Each replicate can be viewed as an in silico patient having the selected number of samples and a distinct cellularity profile across the samples . | Cancer is a genetic disease , driven by DNA mutations . Each tumor is composed of millions of cells with differing genetic profiles that compete with each other for resources in a process similar to Darwinian evolution . We describe a computational framework to model tumor evolution on the cellular level , using next-generation sequencing . The framework is the first to apply a rigorous statistical hypothesis test designed to inform a new search algorithm . Both the test and the algorithm are based on evolutionary principles . The utility of the framework is shown in computer simulations and by automated reconstruction of the cellular evolution underlying murine small cell lung cancers , acute myeloid leukemias and chronic lymophocytic leukemias , from three recent published studies . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [] | 2015 | SubClonal Hierarchy Inference from Somatic Mutations: Automatic Reconstruction of Cancer Evolutionary Trees from Multi-region Next Generation Sequencing |
Mitochondrial dysfunction has been implicated in the pathogenesis of biliary atresia ( BA ) . This study aimed to determine whether a specific mitochondrial DNA haplogroup is implicated in the pathogenesis and prognosis of BA . We determined 40 mitochondrial single nucleotide polymorphisms in 15 major mitochondrial haplogroups by the use of 24-plex PCR and fluorescent beads combined with sequence-specific oligonucleotide probes in 71 patients with BA and in 200 controls in the Taiwanese population of ethnic Chinese background . The haplogroup B4 and E prevalence were significantly lower and higher respectively , in the patients with BA than in the controls ( odds ratios , 0 . 82 [p = 0 . 007] and 7 . 36 [p = 0 . 032] respectively ) in multivariate logistic-regression analysis . The 3-year survival rate with native liver was significantly lower in haplogroup E than the other haplogroups ( P = 0 . 037 ) . A cytoplasmic hybrid ( cybrid ) was obtained from human 143B osteosarcoma cells devoid of mtDNA ( ρ0 cell ) and was fused with specific mtDNA bearing E and B4 haplogroups donated by healthy Taiwanese subjects . Chenodeoxycholic acid treatment resulted in significantly lower free radical production , higher mitochondrial membrane potential , more viable cells , and fewer apoptotic cybrid B4 cells than parental 143B and cybrid E cells . Bile acid treatment resulted in a significantly greater protective mitochondrial reaction with significantly higher mitochondrial DNA copy number and mitofusin 1 and 2 concentrations in cybrid B4 and parental cells than in cybrid E cells . The results of the study suggested that the specific mitochondrial DNA haplogroups B4 and E were not only associated with lower and higher prevalence of BA respectively , in the study population , but also with differential susceptibility to hydrophobic bile acid in the cybrid harboring different haplogroups .
Biliary atresia ( BA ) is characterized by progressive inflammatory cholangiopathy in infancy . The pathogenesis of BA may involve free radical products derived from mitochondria , resulting in significantly increased superoxide dismutase activity [1] . In the early stage of BA , augmented oxidative DNA and mitochondrial DNA ( mtDNA ) damage and apoptotic activities were also observed , which was associated with a decreased mitochondrial copy number [2] . The incidence of BA has been reported to be high in some geographic areas , including French Polynesia and Taiwan [3]–[5] . An interesting finding from a mitochondrial lineage analysis revealed that a specific mitochondrial haplogroup was associated with particular aboriginal races in Polynesia and Taiwan [6] , with a rapid dispersal of maternal lineages from Taiwan approximately 4000 years ago [7]–[9] . Mitochondria are maternally inherited and accumulate mutations faster than does nuclear DNA [6] . An analysis of normal human mitochondrial DNA ( mtDNA ) identified many haplogroup-specific patterns of polymorphisms that have arisen over the last 150 , 000–200 , 000 years [10]–[12] . Studies in the past 10 years have identified an association of mtDNA single-nucleotide polymorphisms with many diseases [13]–[15] . The mtDNA mutation and mtDNA haplogroups are important in understanding the etiology and pathogenesis of human disease [16]–[18] , as mutations of the mitochondrial D-Loop sequence have been demonstrated to be a risk factor for hepatocellular carcinoma ( HCC ) development [18] . In the present study , we identified the importance of the association of haplogroups B4 and E with BA , which confer resistance or susceptibility to hydrophobic bile acid and possibly to the different incidence of BA in different populations .
The demographic data of 71 patients with BA , 52 children and 148 adults as controls , are shown in Table 1 . Except for age , both control groups were devoid of liver diseases and their genetic distributions were almost identical except for one case of haplogroup E in the adult controls . It was therefore put together in the following statistical comparison of haplogroup prevalence with the patients with BA . There were 4 cases of haplogroup E , all received kasai portoenterostomy . However , none survived with native liver . Among them , 3 received liver transplantation at ages 8 months , 1 year 5 months , 1 year 10 months , and the other one died at 3 years 1 month of age without receiving liver transplantation . The 3-year survival rate with native liver was significantly lower in haplogroup E than the other haplogroups ( 25 . 0% vs 48 . 6% , P = 0 . 037 ) . Evaluation of haplogroup distributions revealed that the haplogroup B4 was less prevalent in patients with BA than in the control group ( odds ratio [OR] , 0 . 71; 95% confidence interval [CI] , 0 . 66–0 . 77; p = 0 . 001 ) . In contrast , haplogroup E was more prevalent in BA than in the control group ( OR , 11 . 88; 95% CI , 1 . 31–108 . 16; p = 0 . 018; Table 2 ) . Multivariate logistic-regression analysis of haplogroups associated with BA with adjustment for age and sex , showing the risk for haplogroup B4 was p = 0 . 007 , OR 0 . 82 , 95%CI 0 . 71–0 . 85 and for haplogroup E was p = 0 . 032 , OR 7 . 36 , 95%CI 1 . 22–94 . 6 . The main differences between our mtDNA B4 and E are: T3027C ( 16SrRNA ) , G4491A ( ND2 , V 8 I ) , G7266T ( COI , S 455 A ) , G7598A ( COII , A 5 T ) , A7934G ( COII , I 117 V ) , A8701G ( ATPase6 , T 59 A ) , A10398G ( ND3 , T 114 A ) , C10400T ( ND3 , T 114 A ) , T14577C ( ND6 , I 33 V ) , T16217C ( D-loop ) , detected following a full-length sequencing of haplogroups B4 and E , shown in Table S1 . Before conducting the experiments on different cybrids , we confirmed the absence of Cox-II in 143B ρ0 cells depleted of mtDNA but not in the wild-type 143B and cybrid B4 and E cells ( Figure S1A ) . Organic anion-transporting polypeptides ( OATPs ) are sodium-independent organic anion transporters found in a variety of tissues , including those of the liver , and in the studied 143B cells . OATPs contribute to the transport of bile acids . To assess the responsiveness of the cybrids to bile acid treatment , we identified the bile acid receptor OATP mRNA ( Figure S1B ) and protein expression in the cybrids ( Figures S1C and D ) . In comparison with the wild-type 143B cells without bile acid treatment , the 143B cells and cybrid B4 and E cells treated with bile acid had a significantly decreased mitochondrial membrane potential ( MMP ) . The change was time- and dose-dependent in all the 3 cell lines ( Figures 1A and B ) . MMP was significantly higher in the cybrid B4 cells than in the cybrid E and parental 143B cells ( Figure 1C ) . A flow cytometric measurement showed that the cybrid B4 cells had lower reactive oxygen species ( ROS ) production than the cybrid E and parental cells ( Figure 2A ) . mtDNA copy number usually measures the responsiveness of mitochondria to stress . In this respect , cybrid B4 was unique in that its mtDNA copy number increased significantly compared with those of the parental 143B and cybrid E cells at 24 and 48 h after the bile acid treatment . On the contrary , mtDNA copy number significantly decreased as early as 6 h after the bile acid treatment in the E cybrid cells compared with the parental 143B and B4 cybrid cells ( Figure 2B ) . Hydrophobic bile acids such as chenodeoxycholic acid are generally toxic to the cells with bile acid transporters . To determine apoptosis , 143B cells and cybrid B4 and E cells were treated with different doses of chenodeoxycholic acid for 24 and 48 h . After treatment , annexin V and PI double staining was performed , revealing significantly less apoptosis in the cybrid B4 cells than in the cybrid E and parental cells ( Figure 3A , after 24 h of treatment ) . ( 3- ( 4 , 5-Dimethylthiazol-2-yl ) -2 , 5-diphenyltetrazolium bromide ) , indicated a dose-dependent decrease in viable cells among the 143Bcells , and cybrid B4 and E cells treated with increasing amount of bile acid . However , the cybrid E cells were most vulnerable to the bile acid treatment , having significantly fewer viable cells than the B4 and parental 143B cells at a dose higher than 200 µM for 24 and 48 h after the treatment ( Figure 3B , after 24 h of treatment ) . Mitochondrial fusion is essential for organelle function and cellular homeostasis [19] . To assess mitochondrial fusion , the 143B cells , and cybrid B4 and E cells were treated with 200-µM chenodeoxycholic acid for 6 , 24 , and 48 h and were harvested to determine the mitofusin 1 ( Mfn1 ) and mitofusin 2 ( Mfn2 ) expression levels . As expected , significantly higher Mfn1 and Mfn2 expression levels were found in the cybrid B4 cells than in the cybrid E and parental 143B cells at 24 and 48 h after treatment ( Figures 4A and 4B ) . Our group has reported that cybrids with various mtDNA haplogroups demonstrated differences in oxidative phosphorylation ( OXPHOS ) capacity and viability under oxidative stress , suggesting mtDNA polymorphism contributes to differential tolerance against exogenous insult [20] . As shown in Figure 3B , cybrid B4 had better cell viability than cybrid E after exposure to bile acid . To assess whether the difference was due to different OXPHOS capacity and glycolysis , oxygen consumption rate ( OCR ) and lactate production were examined . Cybrid B4 had higher OCR in which ADP-limited ( state 4 ) , ADP-stimulated ( state 3 ) , and FCCP-induced ( uncoupled ) respiration than did cybrid E , 143B and ρ0 in either absence or presence of bile acid exposure ( Figures 5A and 5B ) . In contrast , following bile acid exposure , cybrid E and parental 143B , but not cybrid B4 and ρ0 , demonstrated a surge of lactate production ( Figure 5C ) . These results indicated that the variations in tolerance against bile acid are correlated with mtDNA polymorphism-associated mitochondrial OXPHOS capacity . The differential viability and respiratory function in the presence of bile acid couple with metabolic shift to glycolysis .
In this study , we found that the mtDNA haplogroups B4 and E were associated with BA , although with entirely different implications . We found that whereas the haplogroup B4 was significantly less likely to occur in the patients with BA , the haplogroup E was more likely to occur in the patients with BA than in the healthy controls . Given that deregulation of mitochondrial function has been implicated in the pathogenesis of BA [1] , [2] , we studied the differential responses of cybrids B4 and E to a hydrophobic bile acid ( chenodeoxycholic acid ) -induced oxidative injury . As expected , we found that cybrid B4 maintained a significantly higher MMP and produced significantly less oxygen free radicals than the cybrid E and parental cells in response to the bile acid treatment . Furthermore , mitochondrial copy number and mitochondrial fusion protein expression , two protective responses of mitochondria to stress , were significantly higher in the cybrid B4 cells than in the cybrid E and parental cells , which was associated with significantly fewer apoptotic cybrid B4 cells than apoptotic cybrid E and parental cells . Overall , these findings suggested a protective effect of the haplogroup B4 from oxidative liver injury , which is at least in part due to inherently better mitochondrial function . Consistent with our findings is a significantly higher evolutionary rate of mtDNA than that of nuclear DNA [21] because mtDNA is more vulnerable to high levels of ROS , lack of protective histones , and limited capacity for DNA repair than nuclear DNA [22] . Mitochondrial dysfunction has been linked to organ failures [23] , [24] , and polymorphisms in the mtDNA are expected to contribute more extensively to functional differences between individuals compared with polymorphisms in nuclear DNA [15] . It is not surprising that many human diseases have been associated with pathogenic mutations in the mtDNA or haplogroup-associated polymorphisms [16] , [25] , [26] . In particular , the sequence polymorphism in the D-loop region of the mtDNA was a risk factor for HCC [18] . The single nucleotide polymorphisms in the D-loop region of mtDNA 16266C/T , 16293A/G , 16299A/G , 16303G/A , 242C/T , 368A/G , and 462C/T minor alleles , were associated with increased risk for alcohol- HCC , and the 523A/del was associated with increased risks of alcohol- HCC and HBV-HCC [18] . Each mammalian cell contains several hundreds to more than a thousand mitochondria , and each organelle harbors 2–10 copies of mtDNA [27] . The mtDNA copy number in a specific cell type is usually precisely maintained but might be subjected to change during cell growth and differentiation [28]–[30] . In our study , the increased mtDNA copy number in the haplogroup B4 was a protective response , which is similar to the protective response of human beings against sepsis either through increased heat generation because of higher electron transport rates or looser coupling [31] , or through ROS production to reduce infection [12] . About 96% of our population group , including the B4 cybrid line used in this study , carry T16217C but only 12% carry the additional A16247G , C16261T control region variants [32] , [33] . The combinations of these variants are specifically identified as a Polynesian motif , which has also been reported in some of the Chinese population [32] . In our study ( Table S1 ) , there are eight mtDNA sequence alterations between B4 and E which cause amino acid changes . Four alterations occurred at subunits of respiratory complex I , three at complex IV and one change at complex V . Previously , mitochondrial complexes I and IV were suggested as being the main site of mitochondrial electron leakage [34] . Thus , it is feasible to hypothesize that amino acid alteration in the subunits of these enzyme complexes could lead to mitochondrial functional change , including a higher oxidative stress in haplogroup E , which is associated with significantly lower MMP , higher ROS production , and thus higher cellular apoptosis . The results also imply less likely occurrence of the haplogroup B4 in the patients with BA than in the haplogroup E . Initial decreases in MMP and subsequent ROS production are a characteristic feature of pre-apoptosis [35]–[38] . In the present study , the bile acid treatment resulted in an initial significant decrease in MMP , which was associated with an excessive intracellular ROS production in the cybrid and parental cells in a time- and dose-dependent manner . The mtDNA-based differences , including different mtDNA and mtRNA levels , mitochondrial protein synthesis , cytochrome oxidase activity and level , normalized oxygen consumption , and mitochondrial inner membrane potential and growth capacity , were also found between cybrids [39] . Such inherited differences in mitochondrial function , including oxidative phosphorylation capacity [40] , can help to explain the differential bioenergetic threshold toward oxidative liver injury between populations with different haplogroups that confer their different susceptibility to BA . Occasional dysfunctional mitochondria can be complemented functionally through mitochondrial fusion . In cells lacking mitochondrial fusion , dysfunctional mitochondria may accumulate , leading to injury-induced cell death [39] . Mfn1 is the main mediator of mitochondrial fusion and homeostasis [41] . High Mfn1 expression levels significantly improved embryo development by increasing ATP levels and MMP while reducing H2O2 generation [41] . Mfn2 is also important in mtDNA transmission to daughter cells by tethering the mitochondrial membrane [42] . Downregulation of Mfn2 was found to induce cellular apoptosis , whereas overexpression of Mfn2 diminished cellular apoptosis [39] . In our study , we found that the cybrid B4 cells had higher Mfn1 and Mfn 2 expression levels than the cybrid E and parental cells in response to the bile acid treatment . The latter may also help to explain why apoptosis was significantly less in the cybrid B4 cells than in cybrid E and parental cells after the bile acid treatment . Kim reported that synthetic bile acids induce apoptosis of osteosarcoma cells through a caspase and mitochondrial pathway [43] . Bile acid increases osteoblast differentiation , and neutralizes the detrimental effects of lithocholic acid in jaundiced patients [44] . In this study , the parental osteosarcoma cells harboring haplogroup X , after bile acid treatment , had different MMP , higher mtDNA copy number , different ROS production , higher cellular apoptosis and lower Mfn1 and Mfn 2 expression levels . We cannot exclude the differential effects of the hapogroup B4 and E from “wild type” as related to their mtDNA haplogroup , as well as to the change of the characteristics of osteosarcoma cells during the process of constructing cybrids . From the findings that the patients with BA harboring haplogroup E had worse prognosis than other haplogroups and had to receive liver transplant earlier or else died , it is reasonable to check haplogroups in every BA case in the future to predict the outcomes as early as possible . In conclusion , the study provides a novel insight into the etiopathogenesis of BA , based on the genetic background of mitochondria . The use of the cybrid models helped to establish the association of haplogroups B4 and E with BA and to clarify the population-based differences in the incidence of BA .
The study was approved by the Ethics and Clinical Research Committee of the Chang Gung Memorial Hospital ( No . 97-2268A3 ) . Seventy-one patients with BA and 200 healthy controls in the Taiwanese population of ethnic Chinese background were recruited from the Kaohsiung Chang Gung Memorial Hospital . BA was confirmed on the basis of results of liver histological examination and intraoperative findings , which included 35 males and 36 females ranging in age from 0 . 1–11 years ( median 2 . 7 ) . All received screening for cytomegalovirus , rubella , herpes infection , but no specific causative factors were identified . As our hospital is the biggest referring center for liver transplantation in Taiwan , there are more patients referred for liver transplant after failed Kasai procedure in the other hospitals than receiving the operation primarily in this hospital . Accordingly , only eighteen patients survived with their native liver , while the remaining 53 patients either survived with liver transplant ( 46 ) or died ( 7 ) . The control group included 52 children and 148 adults . The 52 children were originally recruited for test of reaction to allergens and were proven negative for allergy . There were 25 males and 27 females , ranging in age from 0 . 3 to 16 years ( medium 2 . 0 ) . The adult population comprised 65 males and 83 females , ranging in age from 18 to 60 ( medium 41 ) . They were healthy Taiwanese subjects randomly selected from the health screening center . None of the subjects had a history of BA , cholestatic or autoimmune liver diseases ( Table 1 ) . The genetic background was similar in both control groups and was put together for statistical calculation . Informed written consent was obtained from their legal guardians or the participants for use of their blood in the study . We performed haplogroup analysis , which included 40 mitochondrial single nucleotide polymorphisms in 15 major mitochondrial haplogroups . Genomic DNA was extracted from the whole blood . We used 24 pair primers to perform the gene amplification by multiplex polymerase chain reaction ( PCR ) , which has been described as in Liou et al [32] , [45] . The range of amplicon size was 190–300 base pairs ( bp ) . We used 94 probes for mitochondrial haplogroup determination [32] , [45] . The reactions were then measured by the Luminex100 flow cytometer . The detailed method for genotyping has been described in the protocol by Itoh Y et al [32] , [45] , [46] . Venous blood ( 3 mL ) was collected from each subject into tubes containing 50 mmol/L EDTA ( disodium salt ) , and genomic DNA was isolated by use of a commercial kit ( Genomix , Talent , Trieste , Italy ) [32] . Mitochondrial polymorphisms were determined with sequence-specific oligonucleotide probes by the use of suspension array technology ( Luminex 100; Luminex , Austin , TX ) . The details have been described as in Liou et al [32] . The human mitochondrial single nucleotide polymorphism ( mtSNP ) database provided in the Mitomap website ( http://www . mitomap . org/MITOMAP ) was used for reference . We selected 40 mtSNPs that defined 15 major haplogroups ( A , B , C , D , E , F , G , M7 , M8 , M9 , M10 , M11 , M12 , M13 , N9 ) and their constitutive 13 sub-haplogroups ( B4 , B5 , D4 , D5 , F1 , F2 , F3 , F4 , M7a , M7b , M7c , M8a , N9a ) in this study . The selection was based on the previously constructed phylogenetic trees for Chinese and Japanese populations [32] , [47] , [48] . Human osteosarcoma 143B cells were grown in Dulbecco's modified Eagle's medium ( DMEM ) containing 4 . 5 g/liter of glucose , supplemented with 10% ( v/v ) FBS , 50 units of penicillin/ml , and 50 µg/ml of streptomycin at 37°C in 5% CO2 . For ρ0 cells generation , cells were grown in the presence of 50 ng/ml ethidium bromide in DMEM containing 4 . 5 g/liter of glucose , supplemented with 10% ( v/v ) FBS , 50 units of penicillin/ml , 50 µg/ml of streptomycin , 50 µg/ml of uridine , and 110 µg/ml of sodium pyruvate ( ρ0 medium ) at 37°C in 5% CO2 for three months . The medium was changed every 2–3 days and the cells were re-plated once a week . The absence of mitochondrial DNA was determined by real-time PCR or by Western blot confirmation of low or absent cytochrome c oxidase-II ( Cox-II ) expression [49] . Platelets were used as mtDNA donor . Fusion of platelets from the healthy Taiwanese volunteers with haplogroup B4 and E with 143B ρ0 cells were carried out in the presence of 50% ( wt/vol ) polyethylene glycol 1500 ( Boehringer Mannheim ) . The fusion mixture was recovered for 1 week in ρ0 medium and then cultivated in selection DMEM medium , in which even cybrids with very low Cox-II expression were shown to grow without pyruvate and uridine . Negative control plates consisted of 143B ρ0 cells , which had undergone trans-mitochondrial fusion in the absence of platelets . On days 14–30 after fusion , successful cybrid clones growing in the medium were isolated clonally by limited dilution , while negative control cells died out . The presence of exogenously imported mtDNA , or specific mtDNA haplogroup in the cybrid was confirmed by Luminex100™ system . Human osteosarcoma 143B cells , 143Bρ0 cells and cybrids carrying B4 and E haplogroups were washed once with phosphate-buffered saline ( PBS ) , followed by the addition of 1 mL DMEM containing 0 . 05 mg/mL 3- ( 4 , 5-dimethylthiazol-2-yl ) -2 and 5-diphenyltetrazolium bromide ( MTT; Sigma ) . After incubation at 37°C for 1 h , the media were removed and the formazan crystals in the cells were dissolved in 1 mL DMSO . Cell viability was detected at an OD570 nm using a spectrophotometer . To evaluate ROS levels , cybrid B4 and E cells ( 6×105 cells/flask ) were incubated with different concentrations of chenodeoxycholic acid for different time periods . Cells were then stained with 10 µM of Dichlorofluorescein diacetate for 30 min at 37°C , and then detached with trypsin/EDTA . Cells were collected in 1× PBS , washed twice by centrifugation ( 1500 rpm; 5 min ) , and resuspended in 0 . 5 mL 1× PBS for ROS production . It was measured by flow cytometry utilizing a BD Biosciences FACScan system . ROS production was expressed as mean fluorescence intensity ( MFI ) , which was calculated by CellQuest software . 6×105 cells were plated in six-well plates and allowed to attach for 16–18 h . After being treated with bile acids at different concentration for 24 h or 48 h , the cells were harvested by treatment with trypsin , washed in PBS , and resuspended in 200 ng/mL of Rhodamine 123 ( Invitrogen ) . After incubation for 30 min at 37°C , the cells were washed three times and resuspended in 500 µL of PBS . Cytofluorimetric analysis was performed using a fluorescence-activated cell scanner machine ( BD Biosciences FACScan system ) [2] , [49] . DNA samples were extracted from cybrid cells . The mtDNA copy numbers were measured by a real-time PCR and corrected by simultaneous measurement of the nuclear DNA ( β-actin gene ) . The forward and reverse primers complementary to the nuclear β-actin gene and ND1 gene were as presented before [2] . The threshold cycle number ( Ct ) values of the β-actin gene and the ND1 gene were determined for each individual quantitative PCR run . Relative copy number ( Rc ) = 2ΔCT , where ΔCT was the Ct ß-actin - Ct ND1 [50] . Each measurement was carried out at least three times and normalized in each experiment against a serial dilution of a control DNA sample . The cybrid B4 and , E cells ( 6×105 cells/flask ) were incubated with various concentrations of bile and H2O2 and harvested at 6 , 24 and 48 h respectively . The percentage of cells in different phases of cell cycle was analyzed by flow cytometry after staining with propidium iodide ( PI ) . Methodology is described briefly as follows; cells were fixed overnight with 70% ethanol at 4°C and washed twice with PBS . Then cells were centrifuged 1 , 500 RPM for 5 min and supernatant discarded . The cell pellet was resuspended in 1 ml PI ( 50 µg/ml ) /0 . 05%Triton X-100 staining solution with RNase A ( 0 . 1 mg/ml ) and incubated for 30 min at 37°C . Additionally , an apoptosis assay was performed using an Annexin V-FITC Apoptosis Detection Kit ( BD Biosciences ) . The samples were analyzed on a Becton Dickinson FACScan flow cytometer using the CellQuest software . Annexin V positive/PI negative and Annexin V positive/PI positive cells were defined as necrotic and apoptotic cells respectively . The cells were homogenized in a buffer then centrifuged at 14 , 000 g . Protein ( 40 µg ) from the supernatant of each sample was separated by SDS-PAGE and transferred to polyvinylidene difluoride membranes for electrophoresis . The membranes were blocked in TBST buffer for 1 h at room temperature after which the blots were incubated with a primary cytochrome c oxidase-II ( Cox-II ) , OATP ( organic anion-transporting polypeptide ) , Mfn1 , Mfn2 antibody ( Santa Cruz Biotech . ) , followed by a secondary alkaline phosphatase-conjugated anti-IgG antibody ( 1∶5000; Promega ) . The Western blots were visualized with the Blot AP System ( Promega ) . Alpha-tubulin or b-actin was used as internal control . Detection of oxygen consumption in permeabilized cells was described previously with some modifications [51] , [52] . Oxygen consumption was monitored by a Clark electrode ( Mitocell S200 micro respirometry system; Strathkelvin Instruments , Motherwell , UK ) . Cells ( 100 µL at 5×106 cells/mL ) in a respiratory medium ( 100 mM KCl , 3 mM MgCl2 , 20 mM HEPES , 1 mM EDTA , and 5 mM KH2PO4; pH 7 . 4 ) were permeabilized by digitonin ( optimal concentration 32 . 5 µg/mL as determined by trypan blue staining ) and loaded into a 200 µL MT200 Respirometer chamber , suspended by a fixed-speed solid-state magnetic stirrer inside the chamber and maintained at 37°C by a circulating water bath . Mitochondrial state 4 respiration was determined using 10 mM glutamate/malate as substrates of respiratory complex I in the absence of ADP . Subsequently , mitochondrial state 3 was measured in the presence of 0 . 2 mM ADP . The following uncoupled respiration was determined by the addition of 1 µM carbonyl cyanide 4- ( trifluoromethoxy ) phenylhydrazone ( FCCP ) to the respiratory medium . Oxygen level in medium was expressed in nanomoles of O2 per mL . Oxygen consumption rate was expressed in nanomoles of O2 per minute per 106 cells . Lactate was measured in collected culture medium using the Lactate Colorimetric Assay Kit ( Abcam ) . Frequencies of all mitochondrial haplogroups in the BA patients and controls were tested for independency using Pearson chi-square statistics and Fisher's exact test as appropriate . Cybrid data were expressed as mean ± SE . Comparisons between different cybrids and parental cells were analyzed by the ANOVA with Bonferroni's correction when multiple comparisons were evaluated . We performed multivariate logistic-regression analysis , with BA as a dependent variable and independent variables including age and sex . Patient survival was assessed using the Kaplan–Meier method and compared between groups using the log–rank test . Survival with native liver was analyzed using time from birth until the time of follow-up excluding those who died or received liver transplantation . A p-value <0 . 05 was considered statistically significant . The statistical analyses were performed using the Statistical Package for Social Science ( SPSS , version 12 ) software package . | Mitochondrial dysfunction has been implicated in the pathogenesis of biliary atresia ( BA ) . We determined 40 mitochondrial single nucleotide polymorphisms in different mitochondrial haplogroups in BA patients and controls . The prevalence of haplogroup B4 and E was significantly lower and higher respectively , in the patients with BA than in the controls . The survival rate with native liver was significantly lower in haplogroup E than the other haplogroups . The in vitro study using cybrid cells revealed significantly lower free radical production , higher mitochondrial membrane potential , higher mitochondrial DNA copy number and fewer apoptotic in cybrid B4 cells than cybrid E cells . The study provides a novel insight into the etiopathogenesis and the predictive value of mitochondrial haplogroups in BA . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"haplotypes",
"genomics",
"genetic",
"polymorphism",
"genetic",
"association",
"studies",
"genetic",
"screens",
"genetics",
"population",
"genetics",
"genetic",
"toxicology",
"biology",
"human",
"genetics",
"genetic",
"testing",
"genomic",
"medicine",
"toxicology",
"mitochondrial",
"diseases"
] | 2013 | Associations of Mitochondrial Haplogroups B4 and E with Biliary Atresia and Differential Susceptibility to Hydrophobic Bile Acid |
Dyskeratosis Congenita ( DC ) is a heritable multi-system disorder caused by abnormally short telomeres . Clinically diagnosed by the mucocutaneous symptoms , DC patients are at high risk for bone marrow failure , pulmonary fibrosis , and multiple types of cancers . We have recapitulated the most common DC-causing mutation in the shelterin component TIN2 by introducing a TIN2-R282H mutation into cultured telomerase-positive human cells via a knock-in approach . The resulting heterozygous TIN2-R282H mutation does not perturb occupancy of other shelterin components on telomeres , result in activation of telomeric DNA damage signaling or exhibit other characteristics indicative of a telomere deprotection defect . Using a novel assay that monitors the frequency and extension rate of telomerase activity at individual telomeres , we show instead that telomerase elongates telomeres at a reduced frequency in TIN2-R282H heterozygous cells; this recruitment defect is further corroborated by examining the effect of this mutation on telomerase-telomere co-localization . These observations suggest a direct role for TIN2 in mediating telomere length through telomerase , separable from its role in telomere protection .
The multi-subunit shelterin complexes bind along mammalian telomeres , shielding the natural chromosome ends from engaging the DNA damage signaling and repair machinery [1] . Among the shelterin components , TRF1 and TRF2 bind directly to duplex telomeric repeats [2] , while POT1 binds to the single-stranded regions of telomeres [3] . TPP1 forms a heterodimer with POT1 and enhances the affinity of POT1 for telomeric ssDNA [4] . Depletion of TPP1 or POT1 results in the deregulation of the single-stranded telomeric terminal overhang and the induction of a DNA damage response at telomeres [5–7] . TIN2 directly interacts with TRF1 , TRF2 and TPP1 , assuring structural integrity of the complex [8–10] . Depletion of TIN2 causes profound telomere deprotection phenotypes including destabilization of the shelterin complex , activation of telomeric DNA damage signaling , and increased apoptosis [9 , 11–14] . Increasing evidence suggests that the shelterin complex also regulates access of telomerase to telomeres and hence telomerase action on them . The best evidence for a shelterin-specific role in telomerase regulation comes from analysis of TPP1 , which interacts with the telomerase catalytic subunit through the N-terminal OB-fold domain of TPP1 [15–18] . This interaction is crucial for recruiting telomerase to telomeres , as assessed by co-localization of telomerase RNA to telomeres through in situ hybridization analysis [19] . The TPP1/POT1 heterodimer also promotes telomerase processivity , as demonstrated by an in vitro direct telomerase activity assay [4 , 20] . Notably , mutations in the TPP1 OB-fold domain compromise telomerase-dependent telomere extension but not telomere end protection [18 , 21] , indicating that TPP1 performs a role in telomerase regulation which is distinct from its contribution to chromosome end protection . Whether other shelterin components also directly contribute to telomerase regulation has been less well characterized . Depletion of TIN2 , which associates with TPP1 , leads to reduced levels of TPP1-mediated telomerase association to telomeres [19] , although this result might simply reflect an indirect function for TIN2 as a regulator of telomerase recruitment through anchoring TPP1 at telomeres . Intriguingly , an N-terminally truncated form of TIN2 lacking the TPP1 interaction domain can still induce significant telomerase-dependent telomere extension [8 , 22] , suggestive of a TPP1-independent role for TIN2 in telomerase regulation . An important resource for genetic defects in both telomerase and shelterin has come from Dyskeratosis Congenita ( DC ) patients . DC is an inherited disorder caused by abnormally short telomeres [23] . Clinically diagnosed by the mucocutaneous abnormalities , DC patients are prone to developing bone marrow failure , multiple types of cancers and a spectrum of diseases collectively characterized as “telomere syndromes” [24] . DC-causative mutations have been found in various telomerase ribonucleoprotein components affecting enzymatic activity ( i . e . the telomerase catalytic subunit TERT and the RNA subunit TER ) as well as telomerase biogenesis and trafficking ( i . e . Dyskerin , NHP2 , NOP10 and TCAB1 ) [25–30] . Recently , additional DC-causative mutations have been identified in shelterin components ( i . e . TIN2 and TPP1 ) and other proteins involved in telomere replication ( i . e . RTEL1 and CTC1 ) [31–38] . TIN2 mutations in DC patients correlate with aberrantly shortened telomeres and early onset of DC . Almost all patients reported thus far are heterozygotes , harboring only one mutated allele of TIN2 . The vast majority of the disease-related TIN2 mutations are missense point mutations that cluster in a highly conserved yet uncharacterized region in the TIN2 C-terminus ( a . a . 280–291 ) , outside of the known TRF1 , TRF2 or TPP1 interaction regions [10 , 39] , with R282H being the most frequently observed mutation . Using an ectopic expression system to examine the consequences of TIN2 DC mutations on telomere length in human cells , one recent study reported that overexpression of TIN2 DC mutants caused accelerated telomere shortening in HT1080-derived HTC75 cells , through a telomerase-dependent pathway [40] . However , another study challenged these conclusions and reported that overexpression of wild-type TIN2 and TIN2 DC mutants produced indistinguishable telomere length changes in HT1080 cells [41] . We note that overexpression studies have serious limitations as models to characterize the mechanistic basis for TIN2 dysfunction in DC patients: First , as observed by one of the above studies [40] , ectopic expression of TIN2 increased endogenous TRF1 and TPP1 levels ( both of which have roles in telomere length regulation ) . In contrast , neither TPP1 nor TRF1 accumulates to higher levels in TIN2 DC cells ( under conditions where the mutant TIN2 protein is expressed at endogenous levels ) than in wild-type cells . Second , since TIN2 binds to multiple shelterin proteins but not directly to telomeric DNA , overexpression of TIN2 can potentially sequester other shelterin proteins from telomeres . These side effects complicate the interpretation of TIN2 overexpression studies and may have caused the discrepancy between the overexpression studies . Collectively , these results have left unaddressed whether TIN2 , like TPP1 , has a direct role in telomerase regulation that can be distinguished from its telomere end protection activity . Here we recapitulate TIN2-R282H heterozygosity in cultured telomerase-positive human cells using a zinc finger nuclease mediated knock-in approach to generate a TIN2 DC allele expressed at its normal level from its endogenous locus . Unlike the DC patient-derived cells which have been carrying the TIN2 mutations for years and may have acquired secondary compensatory mutations or epigenetic changes that potentially complicate the interpretation of phenotypes , our approach allows analysis of the immediate telomere phenotype in isogenic human cell clones that differ only in their TIN2 status . Analyses of these cells demonstrate that the TIN2-R282H heterozygosity has no impact on the telomere protection function of TIN2 . Instead , we show that this separation-of-function defect in TIN2 leads to impaired telomerase recruitment , resulting in a reduced frequency of telomerase-mediated telomere extension events . These observations identify a second subunit of shelterin that mediates telomerase function , thereby further extending the premise that shelterin performs a dual role at telomeres .
Zinc finger nuclease ( ZFN ) mediated gene targeting was used to knock a TIN2 DC mutation into the human colon carcinoma cell line HCT116 . HCT116 cells were chosen because they have active telomerase and wild-type shelterin components , they maintain a stable diploid karyotype suitable for gene targeting , and they are intact for most DNA damage-dependent checkpoints [42] . To knock TIN2 DC mutations into HCT116 cells , we designed a pair of ZFNs that specifically recognize unique sequences within TIN2 exon 6 [43] ( S1A and S1B Fig ) . Two donor template constructs were used in the knock-in study: one carries a G to A mutation within exon 6 to introduce a single amino acid change from Arg to His at position 282 ( R282H , one of the most frequently observed TIN2 mutations in DC patients ) ; the other carries the wild-type sequence ( WT ) to generate the isogenic wild-type control cells . Translational silent mutations were introduced into the donor template at the ZFN recognition sites to prevent binding and cleavage of the construct by the ZFN pair . Two knock-in clones heterozygous for the TIN2-R282H mutation ( clone R282H . 1 and R282H . 2 ) and two knock-in clones wild-type for the TIN2 gene ( clone WT . 1 and WT . 2 ) were established ( S1C , S1D , and S1E Fig; see also Materials and Methods for details ) . Two TIN2 splice variants had been previously identified in human cells . The R282H mutation lies in the middle of the sixth exon shared between the splice variants ( S1B and S2A Figs ) . As shown in S2B Fig , the expected TIN2 splice variants were produced in both the HCT116 parental cells and the knock-in clones , and sequencing of the reverse transcription products confirmed that both the mutated allele and the wild-type allele were transcribed into mRNA in the TIN2-R282H heterozygotes ( S2C Fig ) . The functional difference between the two TIN2 splice variants is not yet characterized . In multiple human cell lines , only TIN2S could be detected [14 , 44 , 45] , possibly due to the low abundance of the TIN2L protein . Immunoblotting analysis with an anti-TIN2 antibody raised against an N-terminal epitope of TIN2 ( a . a . 44–58 ) detected one TIN2 protein band ( TIN2S ) of ~42KD in all clones ( S2D Fig ) , consistent with the other reports . TIN2 protein levels in the R282H heterozygous clones were indistinguishable from those in the WT clones and parental HCT116 cells ( S2D Fig ) , demonstrating that the R282H mutation did not significantly change TIN2 protein stability . Levels of the other five shelterin proteins were indistinguishable between TIN2-WT and TIN2-R282H clones as well ( S2E Fig ) . Furthermore , immunoprecipitation analysis showed that there was no change in the interaction between TIN2 and its shelterin binding partner TRF1 , TRF2 or TPP1 in the TIN2-R282H heterozygotes ( S2F Fig ) . To examine the effect of the TIN2-R282H mutation on telomeres in cells with active telomerase , we monitored telomere length of the HCT116 knock-in cells over successive cell divisions . Because telomere length is a heterogeneous trait , sub-clones of human cell lines with the same genotype can display variations in mean telomere length , as shown in Fig 1A and previously observed [46] . Despite differences in initial telomere length , telomeres in both TIN2-R282H heterozygote clones shortened progressively until they reached a mean telomere length slightly above 2kb; as expected , telomeres in TIN2-WT cells maintained stable lengths ( Fig 1A ) . This establishes that a primary consequence of the TIN2-R282H heterozygous mutation is a progressive reduction in telomere length that occurs even in the presence of telomerase . Notably , TIN2-R282H heterozygotes had a similar proliferation rate as TIN2 wild-type cells , even at late PDs ( Fig 1B ) . Furthermore , immunostaining analysis showed that there were no detectable changes of telomeric localization for either TIN2 or other shelterin proteins in TIN2-R282H heterozygotes ( Fig 1C and 1D ) . In particular , we found no evidence for TPP1 delocalization in response to the TIN2-R282H heterozygous mutation ( Fig 1E ) , arguing that the telomere length phenotype conferred by the TIN2-R282H defect was not simply due to an indirect effect on TPP1 delocalization . These initial observations also suggested that telomere protection was not impaired by the TIN2-R282H defect . To assess this more rigorously , we evaluated telomere dysfunction-induced foci ( TIF ) formation in early and late PD knock-in cells by performing immunostaining with an antibody against the DNA damage marker 53BP1 and telomeric fluorescent in situ hybridization ( FISH ) with a telomeric peptide nucleic acid ( PNA ) probe . Although HCT116 cells were fully functional for telomere dysfunction-induced DNA damage signaling ( as indicated by the localization of 53BP1 to telomeres in HCT116 depleted of TRF2 ) , there was no significant increase of TIFs , even in late PD TIN2-R282H heterozygote cells ( PD51 and PD76 ) ( Fig 1F ) . Together these results show that the heterozygous TIN2-R282H mutation does not cause shelterin redistributions or gross deprotection of telomeres . TIN2 knock-in cells were also collected at early and late PDs for FISH analysis of metaphase spreads and examined for telomere abnormalities . No significant differences were found between TIN2-WT and TIN2-R282H cells at early population doublings . By PD51 , the TIN2-R282H heterozygotes had a statistically significant increase in chromosome ends lacking detectable telomeric signals , which presumably reflects the very short telomeres in these cells . Notably , however , we did not observe an increase in fragile telomeres or chromosome end-to-end fusions ( Fig 1G ) , further supporting the conclusion that TIN2-R282H heterozygosity did not lead to chromosome end deprotection . The progressive telomere shortening in TIN2-R282H heterozygotes can be caused by either a defect in the telomerase pathway or by a telomerase-independent process such as increased telomere degradation . To distinguish between these two possibilities , we asked whether combining the TIN2-R282H mutation with a telomerase defect would confer an additive effect on telomere shortening , which would argue that TIN2-R282H mediated its effect on telomere length through a telomerase-independent mechanism . We ectopically overexpressed a dominant-negative form of telomerase catalytic subunit ( DN-hTERT ) [47] in parallel in TIN2-R282H heterozygotes and TIN2-WT cells ( at PD8 ) . DN-hTERT overexpression suppressed telomerase activity to undetectable levels as expected ( Fig 2A ) , and caused progressive telomere shortening in all HCT116 knock-in clones ( Fig 2B and 2C ) . Measurement of telomere length changes between PD8-PD28 cells showed that the expression of DN-hTERT led to similar rates of telomere shortening in TIN2-WT and TIN2-R282H cells ( Fig 2C ) . The above observations led us to consider that the mechanism underlying the progressive telomere shortening in TIN2 heterozygote cells was due to a defect in the telomerase pathway . To address this , we first monitored activity levels of the telomerase enzyme . Quantitative PCR analysis of endogenous telomerase RNA in HCT116 knock-in clones showed that the TIN2-R282H mutation did not cause a significant change in telomerase RNA levels ( S3A Fig ) . Telomerase TRAP analysis also showed that telomerase enzymatic activity in TIN2-R282H heterozygotes was indistinguishable from that in TIN2-WT cells ( S3B Fig ) , indicating that the progressive telomere shortening in TIN2-R282H heterozygotes was not caused by a reduction of the core telomerase enzymatic activity . Since telomerase levels did not appear to be altered , we employed two assays to monitor the extent of telomerase activity ( this sub-section ) or telomerase recruitment ( the next sub-section ) at individual telomeres . To directly measure telomerase activity in vivo , we designed a novel assay to measure telomerase extension events at individual telomeres , by adapting a FISH-based assay which utilizes a telomerase enzyme that adds variant telomeric repeats to telomeres [48] . When the mutant template telomerase RNA , 47A-hTER , is expressed in telomerase-positive cell lines , it assembles with endogenous telomerase catalytic subunit hTERT into active telomerase to direct the incorporation of TTTGGG variant repeats at telomeres [49–51] . FISH with a ( CCCAAA ) 3 PNA probe specifically detects the TTTGGG variant repeats , whereas the ( CCCTAA ) 3 PNA probe specifically detects the canonical TTAGGG telomeric repeats ( Fig 3A ) . In prior experiments with the mutant 47A-hTER telomerase RNA , it was detrimental to cell growth , causing chromosome end-to-end fusions [48 , 50] . However , in these previous experiments , the steady state expression level of 47A-hTER was at a ≥10:1 ratio relative to the endogenous hTER . In contrast , when the 47A-hTER was expressed at low levels ( at ~1:1 ratio relative to the endogenous hTER; Fig 3B ) for a limited number of cell divisions in HCT116 cells , we did not observe either chromosome fusion or growth inhibition effects ( Fig 3A and S4A and S4B Fig ) , even though mutant TTTGGG repeats were added by telomerase to chromosome termini , as assessed by FISH ( Fig 3A ) . The presence of very low levels of a variant telomerase enzyme allows us to quantify two aspects of telomerase activity at individual telomeres: ( i ) the relative frequency of telomere extension events , as determined by counting the fraction of telomeres that had incorporated TTTGGG repeats; and ( ii ) the relative length of extension at individual telomeres , as determined by measuring the fluorescence intensity of TTTGGG repeats at these newly extended telomeres . This provides a powerful assay to directly monitor the activity of telomerase at individual telomeres , at a resolution that has not been previously attainable by other approaches . Expression of 47A-hTER at a 1:1 ratio relative to the endogenous hTER in each of the knock-in clones ( Fig 3B ) led to the assembly of enzymatically active 47A-hTER-containing telomerase at comparable levels in the knock-in clones , as shown by the 47A-hTER-specific TRAP analysis in Fig 3C ( see Materials and Methods and S5 Fig for TRAP assay conditions that distinguish between 47A-hTER telomerase and wild-type telomerase ) . TRAP analysis also revealed that the expression of 47A-hTER caused only ~10% decrease of wild-type telomerase activity in each of the knock-in clones ( Fig 3D ) , suggesting that 47A-hTER titrated away ~10% of hTERT from the endogenous telomerase complex . The reconstitution of only a small proportion of endogenous telomerase into 47A-hTER-containing telomerase in this time frame ( 8 days ) likely reflects the extreme stability and very long half-life reported for telomerase RNA in telomerase-positive human cancer cells [52] . Due to the very low levels of the reconstituted 47A-hTER telomerase , only a small subset of telomeres in the knock-in clones had incorporated TTTGGG variant repeats ( Fig 3E ) . Using this assay , we observed that the fraction of chromosome ends incorporating the variant repeats per metaphase in TIN2-R282H heterozygote cells was significantly less than that in TIN2-WT cells ( Fig 3E ) , indicating that the R282H mutation caused a reduction in the frequency of telomere extension events . Strikingly , although the number of chromosome ends elongated by telomerase ( based on the incorporation of variant TTTGGG repeats ) was reduced in TIN2-R282H heterozygous cells , the amount of TTTGGG repeats added by the 47A-hTER telomerase at individual telomeres was not affected . This was revealed by measuring the fluorescence intensity of the TTTGGG variant repeat tracts at individual termini . As shown in Fig 3F , the distribution of the TTTGGG signal intensity in TIN2-R282H heterozygote cells was indistinguishable from that in TIN2-WT cells , indicating that the lengths of extension by the 47A-hTER telomerase at individual extended telomeres were comparable irrespective of the TIN2 status . These observations argued that the frequency , but not the extension rate , of telomerase extension events was reduced in response to the TIN2-R282H mutation . The reduced frequency of telomere extension by the reconstituted 47A-hTER telomerase in TIN2-R282H heterozygotes suggested that telomerase recruitment to telomeres was compromised by the TIN2-R282H mutation . As a final step in our analysis of telomerase function , we directly examined the recruitment of endogenous wild-type telomerase to telomeres by their co-localization in the HCT116 knock-in cells . To do so , cellular localization of telomerase was monitored by RNA-FISH using established oligonucleotide probes complementary to the telomerase RNA component [18 , 53 , 54] . A mix of three oligonucleotide probes ( ~55 nt long ) were used , each covalently labeled with five red fluorescence dyes , hence marking one telomerase RNA molecule by as many as fifteen fluorescence dyes , significantly amplifying the signal [55] . We carried out immunofluorescence staining against telomeric shelterin proteins TRF1 and TRF2 , followed by FISH for telomerase RNA and analyzed the co-localization between telomerase RNA and telomeres ( Fig 4A ) . Strikingly , we observed that the co-localization between telomerase RNA and telomeres was significantly lower in TIN2-R282H heterozygotes than in TIN2-WT cells ( Fig 4B ) . These observations , combined with those shown in Fig 3 , provide direct evidence that the heterozygous TIN2-R282H mutation impairs recruitment of telomerase to telomeres .
In this study , we have identified a novel function for the TIN2 subunit of shelterin , through the analysis of a separation-of-function TIN2 allele recovered from human Dyskeratosis Congenita patients . Using karyotypically stable , telomerase-positive human HCT116 cells , we have generated knock-in clones heterozygous for the DC-associated TIN2-R282H mutation . Notably , the resulting TIN2-R282H heterozygote cells do not display any characteristics of a telomere end protection defect . Instead , two independent systems for interrogating in vivo telomerase function at individual telomeres—a mutant repeat incorporation assay and the co-localization of telomerase RNA to telomeres—show that the TIN2-R282H mutation impairs telomerase recruitment , resulting in a reduction of the frequency of telomere extension by telomerase . Our model is consistent with a previous report that ectopically expressed TIN2-R282H pulls down less telomerase activity than wild-type TIN2 [40] . Originally defined of its role in chromosome end protection , the shelterin complex was thought to negatively regulate telomere length by sequestering telomeres away from telomerase . Functional characterization of the shelterin subunit TPP1 , however , revealed two surprising roles of TPP1 in promoting telomerase recruitment and telomerase processivity , shifting the view of shelterin as solely an end protection complex which blocks telomerase from acting on telomeres . The work described here further demonstrates that a shelterin-dependent role in promoting telomerase function is not unique to the TPP1 subunit , as revealed by the pronounced telomere shortening in the TIN2-R282H mutant cells . Since TPP1 localization to telomeres is unaffected in TIN2-R282H heterozygotes , this argues that the telomerase recruitment defect of the TIN2-R282H mutant is not conferred through a defect in anchoring TPP1 at telomeres . However , it remains possible that the TIN2-R282H mutation renders TPP1 incompetent for interaction with telomerase . Furthermore , our data show that although telomerase is recruited to telomeres at a reduced frequency in TIN2-R282H heterozygotes , the average length of the extension product at those telomeres that are extended by telomerase is unaffected , suggesting that once telomerase is recruited to telomeres , it is as active as in TIN2-WT cells . Whether telomerase recruitment proceeds through a single coordinated pathway that involves the cooperative behavior of both TIN2 and TPP1 , or through independent contributions by these two shelterin subunits , will be a subject for future investigation . For example , whereas TPP1 may have a direct role in recruitment , TIN2 may modify the conformation of telomeres , making them more accessible to telomerase . Interestingly , TIN2 binds to the heterochromatin protein 1γ ( HP1γ ) through the same region where DC-associated TIN2 mutations cluster . Disrupting TIN2-HP1γ interaction impacts both telomere cohesion and telomere length regulation [22] . HP1γ is required for establishing appropriate sister telomere cohesion and may be involved in shaping the local telomeric chromatin into a more favorable structure for telomerase association . Of great interest is why the telomere maintenance defects in DC patients carrying heterozygous TIN2 mutations are usually worse than in those carrying heterozygous mutations in the telomerase enzymatic components . Studies of telomerase in induced pluripotent stem cells ( iPSCs ) have shown that the expression levels of telomerase catalytic subunit and telomerase RNA were both up-regulated significantly during the induction of pluripotency [53 , 56] . One potential explanation for the more severe form of the disease observed in DC patients carrying TIN2 mutations may be that during early embryonic development , the amount of TIN2 , but not the core telomerase components , is the limiting factor for regulating telomerase activity . If so , this also suggests that the recruitment function of TIN2 may be a more tractable target for inhibition of telomerase activity during oncogenesis . Finally , we point out that our results complement the recent analysis of another TIN2 DC mutation ( TIN2-K280E ) in knock-in mouse system . This mutation was found to confer both telomerase-dependent and -independent telomere shortening ( the exact mechanisms remain to be characterized ) , as well as cause subtle telomere replication problems [41] . Whether the differences between the two studies are due to a difference in the molecular defect ( s ) of the two TIN2 DC alleles ( TIN2-K280E versus TIN2-R282H ) and/or to the differences in telomere maintenance between the two systems ( normal mouse cells versus human cancer cells ) remains to be determined .
The targeting construct ( S1B Fig ) was assembled by combining the following segments through overlapping PCR: a 2 . 7kb genomic fragment containing the human TINF2 gene , 1 . 8kb DNA fragment containing the puromycin N-acetyltransferase expression cassette flanked by loxP sites , and 1 . 8 kb of 3’ flanking DNA of the human TINF2 gene . MluI sites were engineered at the 5’ and 3’ ends of the construct to clone it into the pBluescript SK vector . Site-directed mutagenesis was used to engineer the silent mutations at the ZFN binding region ( 5’-CCATGCCAGACCCTGGGGGGAAGGGCTCTGAAG-3’ to 5’-CCTTGCCAGACACTGGGAGGCAGAGCTCTGAAG-3’ ) . For the R282H targeting construct , site-directed mutagenesis was used to introduce the R282H ( 5’-GAGCGCCCC-3’ to 5’-GAGCACCCC-3’ ) mutation in exon 6 . The ZFN recognition site is ~160bp from the R282H mutation . Full length sequences of the targeting constructs ( ~6 . 3kb ) were confirmed by DNA sequencing before use . TIN2 exon 6-targeting heterodimeric ZFNs ( named T2-X6-L5+R4 ) were expressed in pCMV-FokI ( DA+RV ) plasmid system . The specificity and efficiency of the ZFNs were described in [43] . 2 . 5x105 HCT116 cells were plated in 6-well plate 24 hours before transfection . Cells were transfected with 4μg of linearized donor plasmid and 0 . 5μg of each ZFN encoding plasmid using 10μl of JetPrime transfection reagent ( Polyplus ) . Puromycin selection was applied two days after transfection . Individual colonies were then picked and expanded . Targeted HCT116 cell clones were screened by Southern blotting of NdeI+KpnI digested genomic DNA . 19 out of 768 clones were identified as correctly targeted clones . PCR analysis was performed on correctly targeted clones using the following primers: F1 , 5’-TCTAGCTGGCCGACACTTCAATCT-3’; R1 , 5’-CCTGCTAACCCTTTTAGGCACAGC-3’; R2 , 5’-CTACCGGTGGATGTGGAATGTGTG-3’ . R1 is specific to the unedited allele and R2 is specific to the targeting construct . F1+R1 and F1+R2 PCR products were sequenced to identify clones that contain only the designed change in sequences . PCR was also performed to amplify TIN2 genomic sequences encompassing all coding regions and sequenced to verify no additional mutations were present . To confirm both edited allele and unedited allele of TIN2 gene were transcribed , RT-PCR analysis was conducted on total RNA using the following primers: RT-F1 ( spanning exons 5 and 6 ) , 5’- TGGCTGCTTCCAGAGTGCTCTGTT-3’; RT-R1 , 5’- TGGCTTCCTGGCCCTAGGAGGTAA-3’ . The target sequence for the TIN2 shRNA is 5’- GAATCCTCCTCAGCAACAA-3’ . After the screen procedure , we identified two TIN2-R282H heterozygous clones and four TIN2-WT knock-in clones . All four WT clones maintain stable telomere length over prolonged passaging . Two WT clones that have initial mean telomere lengths comparable to the respective R282H heterozygous clones were then selected for additional telomere maintenance analysis . Nuclear extracts were made using the NE-PER Nuclear and Cytoplasmic Extraction Reagents ( PIERCE ) . Protein concentrations were determined by performing the Bradford assay ( Bio-Rad ) . Samples were suspended with 2x Laemmli sample buffer , resolved with 10% SDS-PAGE , and detected by Western blotting using the following primary antibodies: mouse anti-TIN2 ( Imgenex ) , mouse anti-TPP1 ( Abnova ) , mouse anti-TRF1 ( Genetex ) , mouse anti-TRF2 ( Millipore ) , rabbit anti-Rap1 ( Bethyl ) , and rabbit anti-POT1 ( Abcam ab21382 ) , followed by horseradish peroxidase-conjugated donkey anti-rabbit or anti-mouse IgG ( Jackson ImmunoResearch ) , and visualized by the ECL prime reagent ( GE Healthcare ) . The nuclear protein p84 was detected with a mouse monoclonal anti-p84 antibody ( Genetex ) as loading controls . Intensities of TIN2 bands were quantified by densitometry using the ImageJ software . Intensities of p84 were used to normalize between different samples . Nuclear extracts were diluted 1:2 in TNE buffer containing 50 mM Tris-HCl ( pH 8 . 0 ) , 150 mM NaCl , 2 mM EDTA , 1% NP-40 , and protease inhibitors . The diluted extracts were precleared by incubating with protein G-Sepharose beads ( Sigma-Aldrich ) at 4°C for 30 min . Immunoprecipitation was carried out by incubating the precleared supernatant with a rabbit polyclonal antibody [8] and protein G-Sepharose beads at 4°C overnight . The beads were washed six times with TNE buffer before fractionation on a 10% SDS-PAGE and Western blotting analysis . 10% of the amount used for the immunoprecipitation was fractionated directly on SDS-PAGE as input . 4–5 μg of genomic DNA was digested with RsaI and HinfI , fractionated by 0 . 5% agarose gel , then transferred to a Hybond XL membrane and hybridized to an end-labeled telomeric probe ( CCCTAA ) 4 . Signals were detected by phosphorimaging ( Molecular Dynamics ) . Mean telomere lengths were analyzed by the ImageQuant software . Briefly , the telomere signal intensity over each lane was measured and plotted . The mean telomeric lengths were determined assuming a Gaussian distribution and calculated according to the positions of molecular weight markers run on the same gel . The pHR’CMV lentiviral expression vector system used in this study was provided by Dr . Didier Trono . Telomerase RNA expressing lentiviral vectors contain the wild-type hTER or 47A-hTER cDNA driven by the IU1 promoter and a GFP gene driven by the CMV promoter [50] . The 47A-hTER template sequence is 3’-CAAACCCAAAC-5’ . DN-hTERT- or luciferase- expressing lentiviral vectors contain the DN-hTERT or luciferase cDNA driven by the CMV promoter , followed by an internal ribosome entry site and a hygromycin resistance gene . DN-hTERT contains the D712A and V713I mutations which confer it catalytically inactive [47] . The day before infection , 2x105 cells were seeded on a 10cm plate and allowed to attach overnight . For viral infection , cells were incubated with virus-containing culture media supplemented with 8μg/ml polybrene . For in vivo telomerase function analysis using 47A-hTER , approximately 2 transducing units ( TU ) of lentivirus per cell were used for each infection . Cells were infected with >90% efficiency as indicated by a GFP expressed from the same lentiviral vector . After 24 hours , the virus-containing media was replaced with fresh media . Infected cells were pooled and passaged for subsequent analysis . For overexpression of DN-hTERT and control luciferase protein , approximately 30 TU of lentivirus per cell were used for each infection . For combined immunofluorescence staining-telomere FISH , cells grown on coverslips were fixed with 4% formaldehyde and permeabilized with 0 . 5% NP40 . Immunostaining was carried out by incubating with one of the following primary antibodies: anti-TRF1 ( Genetex ) , anti-TRF2 ( Millipore ) , anti-TPP1 ( Abnova ) , anti-TIN2 ( Imgenex ) , or anti-53BP1 ( BD Transduction Laboratories ) , followed by incubating with secondary antibody conjugated with Alexa Fluor 488 ( Molecular Probes ) . The cells were fixed again with 4% paraformaldehyde and dehydrated by successive incubation in 70% , 95% and 100% ethanol before subjected to telomeric FISH analysis using a TMR-OO-5’- ( CCCTAA ) 3−3’ PNA probe as described previously [57] . DNA was stained by 0 . 1μg /mL DAPI . Coverslips were then mounted onto glass slides in Prolong Gold Antifade Reagent ( Invitrogen ) . Combined immunofluorescence staining-telomerase RNA FISH was carried out as described [58] . Briefly , cells grown on coverslips were fixed with 4% formaldehyde and permeabilized with 0 . 1% NP40 . Immunostaining was performed by incubating with a mix of anti-TRF1 ( Genetex ) and anti-TRF2 ( Millipore ) antibodies to amplify telomere signal , followed by incubating with secondary antibody conjugated with Alexa Fluor 488 ( Molecular Probes ) . The cells were fixed again with 4% paraformaldehyde and dehydrated by successive incubation in 70% , 95% and 100% ethanol . The cells were subsequently rehydrated in 50% formaldehyde in 2XSSC , incubated in prehybridization solution containing 10% dextran sulfate , 50% formamide , 2XSSC , 1mg/ml E . coli tRNA , 1mg/ml RNase-free BSA , 0 . 5mg/ml salmon sperm DNA , and 2mM vanadyl ribonucleoside complexes . Telomerase RNA FISH was performed by adding a mixture of three Cy3-conjugated telomerase RNA probes ( 30ng of each per coverslip ) [54] to the prehybridization solution and incubating at 37°C in a humidified chamber overnight . The cells were then washed sequentially by 50% formamide in 2XSSC at 37°C , 0 . 1% NP40 in 2XSSC , 1XSSC and PBS . DNA was stained by 0 . 2 μg/ml DAPI and the coverslips were mounted onto glass slides in Prolong Gold Antifade Reagent ( Invitrogen ) . Cell images were acquired with a Nikon Ti-U microscope using a 100x objective and collected as a stack of 0 . 2 μm increments in the z-axis . After deconvolution using the AutoQuant X3 software , images were viewed with the Maximal Projection option on the z-axis . All image files were randomly assigned coded names to allow blinded scoring for spots co-localization and fluorescence intensity quantification . Metaphase spreads and telomere fluorescence in situ hybridization was performed as described [57] , using an Alexa488-OO-5’- ( CCCTAA ) 3−3’ and a TMR-OO-5’- ( CCCAAA ) 3−3’ PNA probe ( Panagene ) . Images were acquired with a Nikon Ti-U microscope using a 60x objective . All image files were randomly assigned coded names to allow blinded scoring of variant repeats incorporation and fluorescence intensity . Telomere fluorescence intensity was quantified using the ImageJ software . Telomeric variant repeats signals on metaphase chromosomes were segmented manually and the integrated intensity from every segment was quantified . For each metaphase , the average background intensity was determined and subtracted from individual telomere signals . Total RNA was extracted with the TRIzol reagent ( Invitrogen ) . cDNA was prepared using the High Capacity RNA-to-cDNA kit ( Invitrogen ) . Real-time PCR was performed using the Power SYBR green PCR master mix ( Invitrogen ) , with respective set of primers at 50nM concentration , on a StepOnePlus Real-Time PCR machine . Telomerase RNA levels were normalized against GAPDH mRNA levels . Primer sets used: hTER forward 5’- GGTGGTGGCCATTTTTTGTC-3’ , hTER reverse 5’-CTAGAATGAACGGTGGAAGGC-3’; GAPDH forward 5’-CATGTTCGTCATGGGTGTGAACCA-3’ , GAPDH reverse 5’-ATGGCATGGACTGTGGTCATGAGT-3’ . Unless otherwise specified , telomerase activity was analyzed using the TRAPeze kit ( Millipore ) per manufacturer's directions . The telomeric extension products were separated by 10% TBE-PAGE and visualized by phosphorimaging ( Molecular Dynamics ) . For 47A-hTER or WT-hTER specific TRAP assay , TRAP reaction was carried out as described in [59] except that the return primer 5’-GCGCGGTACCCATACCCATACCCAAACCCA-3’ was used to detect 47A-hTER activity , and the return primer 5’- GCGCGGTACCCTTACCCT TACCCTAACCCT-3’ was used to detect WT-hTER activity . TRAP products intensity in each lane were quantified by the ImageQuant Software and normalized to the respective internal control intensity . | The shelterin complex protects telomeres from being processed by the DNA damage repair machinery , and also regulates telomerase access and activity at telomeres . The only shelterin subunit known to promote telomerase function is TPP1 , which mediates telomerase recruitment to telomeres and stimulates telomerase processivity . Mutations in shelterin components cause Dyskeratosis Congenita ( DC ) and related disease syndromes due to the inability to maintain telomere homeostasis . In this study , we have identified TIN2-R282H , the most common DC-causing mutation in shelterin subunit TIN2 , as a separation-of-function mutant which impairs telomerase recruitment to telomeres , but not chromosome end protection . The telomerase recruitment defect conferred by TIN2-R282H is likely through a mechanism independent of TIN2’s role in anchoring TPP1 at telomeres , since TPP1 localization to telomeres is unaffected by the mutation . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2015 | The Shelterin TIN2 Subunit Mediates Recruitment of Telomerase to Telomeres |
Although the genotype-phenotype map plays a central role both in Quantitative and Evolutionary Genetics , the formalization of a completely general and satisfactory model of genetic effects , particularly accounting for epistasis , remains a theoretical challenge . Here , we use a two-locus genetic system in simulated populations with epistasis to show the convenience of using a recently developed model , NOIA , to perform estimates of genetic effects and the decomposition of the genetic variance that are orthogonal even under deviations from the Hardy-Weinberg proportions . We develop the theory for how to use this model in interval mapping of quantitative trait loci using Halley-Knott regressions , and we analyze a real data set to illustrate the advantage of using this approach in practice . In this example , we show that departures from the Hardy-Weinberg proportions that are expected by sampling alone substantially alter the orthogonal estimates of genetic effects when other statistical models , like F2 or G2A , are used instead of NOIA . Finally , for the first time from real data , we provide estimates of functional genetic effects as sets of effects of natural allele substitutions in a particular genotype , which enriches the debate on the interpretation of genetic effects as implemented both in functional and in statistical models . We also discuss further implementations leading to a completely general genotype-phenotype map .
There is an increasing interest in Quantitative Genetics and Evolutionary Biology to identify genetic effects , and more particularly gene interactions , on a genome-wide scale and to understand its role in the genetic architecture of complex traits [1] , [2] . Genome scans for quantitative trait loci ( QTL ) have proven to be a successful strategy for identifying genetic effects and interactions . Two of the main issues in the development of QTL mapping methods are which models of genetic effects to use and how to test for effects in regions between marker locations . The second issue is important not only for considering the genome as a virtually continuous space where to map the QTL , but also to efficiently analyze incomplete data sets , which are the norm in practice [3] . Lander and Botstein [4] developed the classic interval mapping ( IM ) method , in which they showed how to perform a QTL mapping strategy implemented with the most likely genotypes for the genome regions in between marker locations , given the genotypes at the flanking markers . This method has been extended in several ways [5]–[8] . Albeit the computation of those likelihoods is complex and time demanding , Haley and Knott [9] , ( see also [10] ) provided a convenient approximation of them by means of a simple regression method . Regarding now the first issue mentioned above—the models of genetic effects—the definition of the genetic effects in Haley and Knott's [9] regression ( hereafter HKR ) comes from a model that has been extensively used in Quantitative Genetics , the F∞ model [11] , [12] . However , other models of genetic effects have recently been shown to be more appropriate in QTL mapping . The genetic effects depend not only on the genotypic values but also on the genotype frequencies of the analyzed population ( e . g . [13]–[16] ) . By taking into account these frequencies , it is possible to build orthogonal models that are convenient for several reasons [13]–[19] . First , orthogonal estimates do not change in reduced models , which considerably facilitates model selection for finding the genetic architecture of traits . Second , the estimates of genetic effects obtained by orthogonal models are meaningful in the population under study—they provide the effects of allele substitutions in that population . Third , they directly lead to a proper , orthogonal decomposition of the genetic variance from which to compute important measures , like the heritability of that trait in that population . The statistical properties of HKR could therefore be improved by implementing it with a genetic model that is orthogonal for any possible genotype frequencies in the population under study . The statistical formulation of the recently developed NOIA ( Natural and Orthogonal InterActions ) model of genetic effects is orthogonal in situations where previous models are not—for departures from the Hardy-Weinberg proportions ( HWP ) at any number of loci—and it is therefore more appropriate choice for estimating genetic effects from data in genetic mapping [16] . Furthermore , a novel feature of NOIA is its implementation to transform the genetic effects estimated in the population under study , in two ways . First , they can be transformed into how they would look like in a population with different genotype frequencies at each locus , like an ideal F2 population or into an outbred population of interest . Second , using the functional formulation of NOIA , it is possible also to express the genetic effects as effects of allele substitutions from reference individual genotypes—instead of from population means like in the statistical formulation . In other words , starting from the orthogonal genetic effects of a population or sample under study , which are the ideal ones for performing model selection and have a particular meaning , NOIA enables us to obtain the values of the genetic effects that are associated to other desired meanings and are useful , therefore , to inspect different aspects of the evolution of a population , or selective breeding for increasing or decreasing a trait values . Our motivation for this communication is to show how to use models of genetic effects to obtain estimates of genetic effects from data that have the desired meaning of any particular scientific purpose . To this end we first inspect how much of a difference it makes to use the classical models for ideal populations , such as ideal F2 populations , to compute genetic effects in a non-ideal situation , under departures from the HWP . We address this issue by generating simulated populations that depart from the HWP in several degrees and analyzing them with NOIA and other models . We quantify the deviances from orthogonal estimates due to using models that assume ideal conditions in the populations under study , thus showing the practical convenience of using the NOIA model for performing real estimates of genetic effects in QTL experiments . Second , we develop an implementation of NOIA with HKR , allowing it for immediate practical use and illustrate its performance using an example with real data . By this example we provide estimates of genetic effects with different meanings and , for the first time , functional estimates of genetic effects—using an individual genotype as reference—from a real data set . We discuss on how this feature opens new possibilities of using real data to analyze important topics in Evolutionary Genetics .
Figure 1 shows the results of estimating , with three different models ( NOIA , G2A and F2 ) , the genetic effects of a two-locus and two-allele genetic system ( Table 1 ) in nine simulated populations under linkage equilibrium ( LE ) with various degrees of departure from the HWP ( see Methods ) . The eight genetic effects plus the population mean in the only model that is orthogonal in all simulated populations—the statistical formulation of the NOIA model—respond to the increasing departures from HWP in three groups . The first and most influenced group contains the three genetic effects involving the additive effect of the locus affected by departures from HWP , αA , αα , and αδ . These genetic effects increase substantially with increasing departures from HWP and are doubled when the homozygote A2A2 is almost completely absent . The second group contains the reference point—the mean of the population , μ—and the single locus effects of locus B ( the one at HWP ) , αB and δB . The estimates in this group decreased with increasing departures from HWP . The third group contains the remaining three genetic effects , δA , δα and δδ , whose estimates are not affected by departures from HWP at locus B . The genetic effects measured by the G2A model show the same qualitative behavior described above for NOIA ( i . e . also responds in three distinct groups ) , but are quantitatively different . The reason for this is that G2A can adapt the measurements to the changes in the allele frequencies of the population , but not to the precise departures of the genotype frequencies from the HWP . The genetic estimates obtained using the F2 model always give the same values independently of the genetic constitution of the population . The F2 thus fails to capture the effects of departures from HWP at all . Thus , unless when the studied population is an ideal F2 ( and the deviances from HWP are zero , see Figure 1 ) , the estimate of the population mean from G2A and F2 is biased and the genetic estimates do not reflect the average effects of allele substitutions in the population under study . Those deviations become more severe as the departure from HWP increases ( Figure 1 ) . Figure 2 shows the variance component estimates obtained in the nine simulated populations , which were obtained by computing the variance over the individuals of the sample population of the correspondent genetic effects ( additive effect at locus A , additive effect at locus B , etc ) . For orthogonal models , the sum of the three components of variance gives the total genetic variance—which in this case equals the phenotypic variance , since there is no environmental variance in the simulated populations . Here , this is only observed for the variances computed using NOIA . The other two models are not orthogonal in the populations under study ( except in the ideal F2 population , where the three models coincide ) , and thus there exist covariances between the genetic effects that would need to be accounted for to obtain the true genetic variance of the population [20] . The decomposition of the genetic variance made by the G2A and F2 models is , thus , non-orthogonal . The G2A leads to a greater departure form an orthogonal decomposition of variance than the F2 model by the particular kind of departures from HWP simulated here . Both the G2A and F2 models underestimate the additive variance and therefore also the heritability of the trait in the simulated populations . For illustrating the advantage of using NOIA for analyzing experimental data , we reanalyze a two-locus ( A and B ) genetic system with epistasis affecting growth rate in an F2 cross between Red junglefowl and White leghorn layer chickens [21] . The two loci are on different chromosomes , thus avoiding linkage disequilibrium ( LD ) . Locus A departs significantly from the HWP when considered alone , but not when correcting for multiple testing ( see Methods ) . Table 2 shows the genetic effects and the components of variance for this two-locus system using several models of genetic effects—NOIA , G2A , F2 and F∞ . As explained in the previous subsection , NOIA is orthogonal under departures from the HWP , whereas the other models are not . The F∞ model deviates severely from the estimates obtained by NOIA . Deviations are expected since the F∞ model is non-orthogonal even in an ideal F2 population with no deviations from the expected frequencies due to sampling errors . The F2 and G2A models , on the other hand , would be orthogonal under ideal circumstances and the observed deviations from orthogonality of those models when analyzing these experimental data are due to sampling ( as explained above ) . Table 2 shows that the estimates obtained using F2 and G2A differ substantially from these of NOIA ( up to 18/42% for the G2A and 53/138% for the F2 model , for the genetic effects/variance component estimates ) . This example with real data , thus , shows that it makes a substantial improvement to use NOIA to compute genetic effects and variance decomposition in QTL mapping experiments over the classical models of genetic effects designed to fit ideal experimental situations . From the statistical estimates in Table 2 , we have computed functional estimates of genetic effects using an analogous expression to ( S6 ) , shown in Text S1 , derived by Álvarez-Castro and Carlborg [16] . The variances of the statistical estimates can also be transformed to give the variances of the functional estimates using ( 6 ) , as derived in the Methods section . Choosing “A1A1B1B1” as reference genotype , the estimates of functional genetic effects , and the standard deviations associated to these estimates , are shown in Table 3 . Whereas statistical genetic effects describe the average effects of allele substitutions in a population , functional genetic effects describe the genotype-phenotype map as a series of allele substitutions performed in the genotype of a particular—reference—individual genotype [16] , [22] , in this case the genotype of the Red junglefowl , “A1A1B1B1” . To illustrate the usefulness of these functional genetic effects for understanding how epistatic effects can contribute to phenotype change , we consider the role of this QTL pair in increasing the growth rate in the Red junglefowl . For simplicity , we assume hereafter that A and B are the only two loci affecting growth rate . From the marginal genetic effects in Table 3 , it can be deduced that the White leghorn layer allele at locus A slightly increases the phenotype whereas the White leghorn allele at locus B actually decreases it , when considered in homozygotes . However , the dominance effects are positive and have a higher absolute value than the additive effects . Therefore , if one White leghorn layer allele appeared by mutation in a Red junglefowl population at any of the two loci , A or B , it would be maintained at a certain frequency because of balancing selection—superiority of the heterozygote—but it would neither disappear nor reach fixation . This suggests that one mutation could be present at some frequency in the population when the second one appeared . For analyzing what would happen if eventually the two mutations were present at the same time in the population , we have to consider also the interaction effects . The double homozygote for White leghorn layer allele increases the phenotype with roughly forty grams ( four times aa , in Table 3 as it can be deduced from G = S⋅E , with the reference of R = G1111 ) , relative to the expected value without epistasis , which is a decrease in roughly 20 grams from the Red junglefowl . In total , this makes the phenotype of the White leghorn layer 20 grams higher than the Red junglefowl . However , for inspecting if this results support the White leghorn layer alleles being likely to reach fixation we also need to consider the phenotypes of the heterozygotes . Interactions involving dominance in locus B are all negative , thus favoring the fixation of the White leghorn layer allele , B2 . The role of allele A2 is not as obvious , since da is positive . The genotypic value of “A1A2B2B2” is roughly 30 grams higher than the Red junglefowl ( computed again from Table 3 and G = S⋅E ) and ten grams higher than the pure White leghorn layer . The expected , therefore , would be that the two alleles segregate at locus A . The standard deviations of the estimates are however rather large and thus do not rule out the possibility of fixation of the White leghorn layer allele at locus A .
The statistical formulation of NOIA is orthogonal under random deviations from ideal experimental populations and outbreeding pedigrees [16] . Therefore , NOIA can provide meaningful estimates of genetic effects—as allele substitutions made in the population or sample under study—and a proper decomposition of the genetic variance under those circumstances . In this article , we illustrate the practical implications of these achievements for estimation of genetic effects and QTL analysis in two ways . First , we simulated a two-locus genetic system under departure from the HWP affecting one of the loci underlying the trait under study . This scenario can have a biological origin or be due to sampling alone and it is commonly occurring in experimental data both from natural and experimental populations , such as for the QTL pair we have studied ( see below ) . We therefore deemed it relevant to test the performance of NOIA in practice—by assessing how departures from HWP cause other models to deviate from the orthogonal values . Our results show that departures from HWP substantially affect both the genetic effects and the decomposition of variance . The cause for this is that epistasis makes the genetic effects dependent on the genetic background , which is different under different degrees of departures from HWP . NOIA can capture the proper , orthogonal genetic effects , and thus also their orthogonal variances , in the simulated populations whereas the deviances from these values due to using the other—nonorthogonal—models increases with the departures from HWP . Second , we used experimental data on epistatic QTL from a previously published study [21] to explore how much of a difference it makes to use NOIA instead of previous statistical models , when departures from HWP are not larger than expected by sampling . Even though the population we studied was rather large ( approximately 800 individuals ) , the random deviations from the HWP in this set of available individuals cause considerable differences in the estimates of genetic effects performed with models that would be orthogonal in totally ideal situations , as compared to the estimates obtained using NOIA . These differences become even more noteworthy for the components of variance estimated using the different models . These values influence consequential quantities , like the heritability of one trait , which may be needed for instance for performing artificial selection at the available sample of individuals . Orthogonal models are also important for finding the genetic architecture of traits—albeit this has not been our focus in this communication . In principle , when testing the effect of a particular locus or set of loci in a QTL analysis , the choice of the model of genetic effects to use does not matter . However , it does matter when it comes to compare which of several putative sets of loci is the most likely genetic architecture underlying the trait , i . e . , when performing model selection in QTL analysis . This is so because orthogonal models have the convenient property that the estimates and their variances remain the same when considering reduced models , which facilitates model selection strategies [19] . After model selection and the estimation of genetic effects have been properly carried out using an orthogonal model , the obtained estimates provide the effects of allele substitutions in the sample of individuals used in the study , and the decomposition of variance is also the appropriate one in that particular sample of individuals . The NOIA model provides convenient tools for transforming those estimates into the ones with any other desired meaning , like the orthogonal estimates and the decomposition of variance in a different population [16] . This is useful to compare results from QTL studies performed in different populations , and to use the results obtained with one orthogonal model in one population to study the evolution of the same trait in a different population . One example of the previous is removing the characteristics of the data that are not supposed to be properties of a target population from the estimates . The departures from HWP of the experimental data we dealt with in this article are in fact supposed to be only due to sampling , instead of being caused by real Hardy-Weinberg disequilibrium in the F2 population . If we were interested in the genetic effects or in the decomposition of variance of the ideal F2 as a target population—in which the departures from HWP are absent—we could use the transformation tool of NOIA to obtain ( from the original estimates with the reference of the mean of the sample population ) the ones with the reference of the mean of an ideal F2 population . Further , as illustrated in the example with real data , it is possible to transform statistical estimates of genetic effects into functional ones , using a particular reference genotype . Another situation in which these transformations are valuable is , for instance , in a three-locus genetic system with pairwise epistasis . In this case , NOIA would easily permit to consider only the significant genetic effects and to re-compute the genotypic values only from the significant genetic effects ( assuming the non-significant third-order interactions to be zero ) . Statistical models of genetic effects are necessary for QTL analysis and for performing orthogonal decompositions of the genetic variance in populations . Functional models of genetic effects , on the other hand , are convenient—especially in the presence of epistasis—for studying evolutionary properties of the populations such us adaptation in the presence of drift and speciation ( see e . g . [23] , [24] ) . NOIA is the first model framework that successfully unifies functional and statistical modeling of genetic effects [16] . This enables researchers to feed models of functional genetic effects , so far mainly used in simulation studies ( see e . g . [2] , [24] ) , with real data obtained using statistical models in QTL mapping experiments . Here , we have actually transformed statistical genetic effects , obtained from real data of an F2 experimental population , into functional genetic effects as allele substitutions performed from a reference individual . Concerning these functional estimates of genetic effects , we have shown in the previous section how they can improve the understanding of the genetic system by inspecting a two-locus model obtained from real data . Notice that when changing the reference of the model , the genetic effects can change their magnitudes and even their signs ( see Tables 2 and 3 ) . Therefore , for reaching the kind of conclusions we obtain above for the evolution of a population from an ancestral genotype “A1A1B1B1” , the genetic effects have to be described with a model that uses that particular genotype as reference point . Those are the only ones that are meaningful for analyzing the problem under consideration . The computation of genetic effects using NOIA in the example with real data required the use of the theory developed in this article , the implementation of the model to handle missing data ( 1 ) . When performing IM for searching for the positions and estimates of genetic effects in QTL mapping experiments , missing data occurs at two levels . First , the genotype of the QTL located in a marker interval is not known and needs to be estimated from the observed flanking marker genotypes . Second , in most experimental datasets there are missing genotypes for many genetic markers that can be imputed from genotypes at closely linked informative markers . Thus , the implementation of HKR with NOIA enables us to perform IM with a regression method and using a model of genetic effects that is orthogonal regardless of how far the available data is from the HWP . The HKR has been assessed as a good approximation of IM when dense marker maps are available and missing data are few and random [25] , [26] , but some disadvantages of this method have also been reported . The residual variance of the HKR has been found to be biased , as first pointed out by Xu [27] . Kao [26] further characterized that bias and found it to be noticeable under LD or strong epistasis . Nevertheless , even in those cases , the estimated genetic effects themselves are not biased [26] . Feenstra et al . [25] have developed a new method , the estimating equation method , which reduces the reported bias of the HKR and is therefore more suitable in the cases when it has proven to be strongly biased . However , the traditional HKR is still popular and convenient mainly due to its dramatic advantage in computational time [25] , and this is why in this study we have chosen this method for implementing NOIA for IM . Models of genetic effects need to be further generalized . Two important cases that need to be accounted for are multiple-alleles and LD , which have been addressed in several recent publications dealing with statistical models of genetic effects . Yang [18] has developed a model to test the importance of LD in QTL data , by designing a component of variance due to LD . This statistical model , like the statistical formulation of NOIA , actually accounts for departures from HWP , although it is restricted to the two-locus case . Wang and Zeng [20] have developed a statistical model with multiple alleles in which they also test the importance of LD , in this case by computing all the covariances between the components of variance , due to LD . It is , however , restricted to HWP . Mao et al . [28] have developed a model to account for LD when computing genetic effects in a two-locus model specially designed for single nucleotide polymorphisms . The desired situation , which we are currently aiming toward is to consider all the different departures from ideal situations gathered under the umbrella of a general formal framework of genetic effects .
We use a simulated numerical example to show how departures from the HWP affect the estimates of genetic effects in several models of genetic effects . We simulate a trait controlled by two biallelic loci , A and B , generating several populations with the second locus affected by departures from the HWP in several degrees . The genotype-phenotype map corresponds to the phenotype mean of the population and all the genetic effects being equal to one in an ideal F2 population ( Table 1 ) . We first constructed data for an ideal F2 population of 800 individuals in strict HWP and LE . From this population we subsequently removed 24 A2A2 individuals and added eight A1A1 and 16 A1A2 individuals in a balanced way , without affecting the population size , the frequencies at locus B , the proportion of A1A1 versus A1A2 individuals or LE . Only deviations from the HWP against the A2A2 homozygote were introduced in the data . We repeated this procedure eight times in total and saved each population data , until only eight A2A2 individuals remained . We measured the departures from HWP in these populations by computing the percentage of reduction of A2A2 individuals relative to A1A1 , which of course was zero in the ideal F2 population we started from . We analyzed the simulated data by computing the genetic effects of the system using three models: NOIA , G2A and F2 . The F2 model , described in Text S1 , is constructed for F2 populations , although it is only orthogonal in ideal F2 populations with the genotypic frequencies being exactly ¼ , ½ , ¼ . The NOIA model is as described in Text S1 . The G2A model [19] accounts for any gene frequencies of—and it is orthogonal at—populations under exact HWP . Álvarez-Castro and Carlborg [16] obtained it as a particular case of NOIA by constraining ( S5 ) , in Text S1 , to HWP:where p is the frequency of allele A1 . The genetic effects were computed for each individual genotype using the genetic-effects design matrices and the estimates of genetic effects from each of the three models , which produced different outcomes . The additive , dominance and interaction variances were obtained as the correspondent sums of the variances of each genetic effect ( for instance , the sum of the variances of the additive effects of each of the loci gives the additive variance ) . We recall the required theory behind the HKR and NOIA in Text S1 . Here we extend the NOIA model to IM with HKR . We do this by implementing the genetic-effects design matrix of the statistical formulation of NOIA , SS ( S5 ) , in the HKR method , as we do with the F2 model in Text S1 . The original genotype frequencies p11 , p12 and p22 in the NOIA statistical formulation ( S5 ) are the exact genotype frequencies at the considered loci . In the HKR , the genotype frequencies are not known , but can be estimated as:where N is the number of individuals in the population under study . We implement this model in the general expression of the HKR ( S4 ) , in Text S1 , and obtain:Let G* be the column-vector of observed phenotypes , G*k , k = 1 , … , N , ε the corresponding vector of errors , and Z , which is an N×3-matrix whose rows are the vectors ωk ( S4 ) . With this notation , the general expression of regression ( S4 ) is: ( 1 ) This has a straightforward extension to several loci with LE . The SS matrix and the E vector can be extended as in Álvarez-Castro and Carlborg [16] . The Z matrix can be extended as the row-wise Kronecker product of the matrices of the single loci , also as in Álvarez-Castro and Carlborg [16] , albeit in that article the matrix accounted for only complete marker information , instead of for IM with HKR , or for missing data probabilities . For instance , for a two-locus ( A and B ) case , the ZAB matrix is an N×9-matrix that is built as: Carlborg et al . [21] identified 10 genome-wide significant QTL for growth rate in chicken from eight to 46 days of age in an F2 intercross of roughly 800 individuals between one Red junglefowl male and three White leghorn females . A simultaneous two-dimensional genome scan was performed to identify pairs of interacting loci regardless of whether their marginal effects were significant or not . We have studied in more detail one of the detected pairs involving QTL on chromosome 2 ( 486 cM ) and 3 ( 117 cM ) , hereafter loci A and B respectively . This pair was selected for a number of reasons . First , these loci interact epistatically , in spite of showing no significant marginal effects in the studied population . Second , since they are located in different chromosomes , there is no physical linkage between them . Third , the genotype frequencies at locus A depart significantly from the HWP ( p<0 . 05 ) when considered independently , but the departure is not significant after applying multiple testing correction accounting for the rest of the detected QTL . Thus , locus A is an example of the departure of the HWP that is expected in QTL experiments just due to sampling . The level of departure from the HWP for the evaluated pair roughly equals the 30% deviation in Figures 1 and 2 . We have computed the genetic effects of the epistatic pair involving loci A and B , using several models of genetic effects . First we used the F∞ model , which was the one also used by Carlborg et al . [21] as it was the model originally implemented in HKR [9] , [29] . Second , the F2 model , which was designed for F2 populations . Third , the G2A model , which can account for departures of the gene frequencies from ½ , and finally the statistical formulation of NOIA , which can adapt to the genotype frequencies of the sample used for the estimation of QTL effects . In these analysis we have made use of the theory developed in this article: the implementation of HKR with NOIA . These developments enable us to deal both with missing data and with the estimation of genetic effects of positions inside the marker intervals . Álvarez-Castro and Carlborg [16] have shown how to transform genetic effects obtained using an orthogonal-statistical model in one population , into statistical genetic effects at any other population or into functional genetic effects from any reference individual . In each of these two cases , the transformation is done as in expression ( S6 ) , in Text S1 , using the S matrix—the genetic-effect design matrix—of the orthogonal system , G = S1⋅E1 , and the inverse of the S matrix in the new system , G = S2⋅E2: ( 2 ) Let ( 3 ) be the transformation matrix . From ( 2 ) and ( 3 ) , the estimates in E1 can be expressed as functions of the estimates in E2 as: ( 4 ) where the letters and their superindexes indicate the vector , or matrix , they are scalars of and the subindexes indicate the position of the scalars inside the vectors or matrices . From ( 2 ) , the variances of the estimates E2 , can be computed from the ones in E1 as: ( 5 ) Now for obtaining the vector of variances of the estimates E2 , V2 , from the vector of variances of the estimates E1 , V1 , we just rewrite ( 3 ) in algebraic notation as: ( 6 ) where the open circle stands for the Hadamard product—giving the matrix whose scalars are the product of the scalars at the same position in the original matrices . | The rediscovery of Mendel's laws of inheritance of genetic factors gave rise to the research field of Genetics at the very beginning of the last century . The idea of traits being determined by the effects of inherited genes is thus the conceptual core of Genetics . After more than one century , however , we still lack a completely general mathematical description of how genes can control traits . Such descriptions are called genotype-phenotype maps , or models of genetic effects , and they become particularly cumbersome in the presence of interaction among genes , also referred to as epistasis . The models of genetic effects are necessary for unraveling the genetic architecture of traits—finding the genes underlying them and obtaining estimates of their individual effects and interactions—and for meaningfully using that information to investigate their evolution and to improve response to selection in traits of economical importance . Here , we illustrate the convenience of using a recently developed model of genetic effects with arbitrary epistasis , NOIA , to inspect the genetic architecture of traits . We implement NOIA for practical use with a regression method and exemplify that theory with a real dataset . Further , we discuss the state of the art of genetic modeling and the future perspectives of this subject . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"genetics",
"and",
"genomics/complex",
"traits",
"evolutionary",
"biology/animal",
"genetics",
"genetics",
"and",
"genomics/gene",
"discovery",
"evolutionary",
"biology/bioinformatics"
] | 2008 | How To Perform Meaningful Estimates of Genetic Effects |
In the nematode Caenorhabditis elegans , different small RNA-dependent gene silencing mechanisms act in the germline to initiate transgenerational gene silencing . Piwi-interacting RNAs ( piRNAs ) can initiate transposon and gene silencing by acting upstream of endogenous short interfering RNAs ( siRNAs ) , which engage a nuclear RNA interference ( RNAi ) pathway to trigger transcriptional gene silencing . Once gene silencing has been established , it can be stably maintained over multiple generations without the requirement of the initial trigger and is also referred to as RNAe or paramutation . This heritable silencing depends on the integrity of the nuclear RNAi pathway . However , the exact mechanism by which silencing is maintained across generations is not understood . Here we demonstrate that silencing of piRNA targets involves the production of two distinct classes of small RNAs with different genetic requirements . The first class , secondary siRNAs , are localized close to the direct target site for piRNAs . Nuclear import of the secondary siRNAs by the Argonaute HRDE-1 leads to the production of a distinct class of small RNAs that map throughout the transcript , which we term tertiary siRNAs . Both classes of small RNAs are necessary for full repression of the target gene and can be maintained independently of the initial piRNA trigger . Consistently , we observed a form of paramutation associated with tertiary siRNAs . Once paramutated , a tertiary siRNA generating allele confers dominant silencing in the progeny regardless of its own transmission , suggesting germline-transmitted siRNAs are sufficient for multigenerational silencing . This work uncovers a multi-step siRNA amplification pathway that promotes germline integrity via epigenetic silencing of endogenous and invading genetic elements . In addition , the same pathway can be engaged in environmentally induced heritable gene silencing and could therefore promote the inheritance of acquired traits .
Transgenerational epigenetic inheritance is the transmission of information from parent to offspring via the gametes by means other than the primary sequence of DNA [1] . A number of recent examples have highlighted the possibility for epigenetic inheritance of acquired traits in animals . This resurrection of Larmarckism is poised to rewrite the textbook on heredity [2] , but a clearer understanding of the molecular mechanisms involved is needed . In the nematode Caenorhabditis elegans , epigenetic inheritance of acquired traits can be triggered by exposure to environmental RNA . In these cases , gene silencing mediated by the RNA interference ( RNAi ) pathway is maintained for multiple generations independently of the initial trigger [3–5] . RNAi is thought to have evolved as a system of surveillance for harmful nucleic acids , and acts to eliminate parasitic nucleic acids such as those derived from viral infection . As such , this epigenetic inheritance pathway may provide in inherited ‘immunity’ against viral infection , or other harmful environmentally derived RNAs . Small RNA pathways related to RNAi also play an important role in defending the genome from endogenous parasitic nucleic acids , such as transposons , retrotransposons , and endogenous retroviruses . Silencing of these elements in germline cells is essential to ensure genome stability , and faithful transmission of genetic material to the next generation [6] . Piwi interacting small RNAs ( piRNAs ) are a class of germline-specific small RNAs that are required for transposon silencing , genome stability and fertility in animals , and have been linked to epigenetic inheritance in a number of species , including the mouse , Drosophila and C . elegans [7] . In this case , epigenetic inheritance may analogously provide an inherited immunity to the potentially harmful nucleic acids harboured within animal genomes . Piwi proteins are a subclass of the Argonaute family of proteins ( Ago ) , a central component of small RNA pathways . Ago proteins physically associate with small , 21–35 nucleotide-long RNAs thereby forming a complex whose specificity is determined by base pairing interactions between the small RNA and target nucleic acids . This typically leads to repression of gene expression , which can occur by various mechanisms [8] . The C . elegans Piwi homologue ( PRG-1 ) , associates with piRNAs ( also termed 21U-RNAs in C . elegans for their characteristic length and 5′ nucleotide ) and directs silencing of transposons and some endogenous genes ( “piRNA targets” ) in the germline , and is important for fertility [9 , 10] . PRG-1 triggers silencing of piRNA targets via a small RNA amplification pathway [9] . Target recognition occurs in the cytoplasm by base-pairing between a piRNA and target mRNA and prompts the generation of secondary siRNAs ( termed 22G-RNAs for their characteristic length and 5′ nucleotide ) antisense to the target mRNA . This process involves the RNA dependent RNA polymerases ( RDRPs ) RRF-1 and EGO-1 , the Dicer-related helicase DRH-3 and a number of proteins in the Mutator ( Mut ) class ( including MUT-2 , -7 , -16 ) [11] , and is thought to take place in perinuclear structures in the germline termed Mutator foci [12] . Efficient silencing of piRNA targets by 22G-RNAs requires the nuclear RNAi pathway , consisiting of the nuclear RNAi factors NRDE-1 , -2 and -4 and the germline-specific nuclear Argonaute HRDE-1 , which binds to 22G-RNAs [3 , 13] . Nuclear RNAi is thought to inhibit gene expression by interfering with transcription , possibly at the elongation stage [14] and may do so by promoting a chromatin state associated with gene silencing [15] . Consistent with a role for repressive histone modifications , efficient silencing of piRNA targets requires the HP1 orthologue HPL-2 , and the putative histone methyltransferases SET-32 and SET-25 [3] . Importantly , although PRG-1 is required to initiate biogenesis of 22G-RNAs antisense to piRNA targets , both 22G-RNAs and target gene silencing can be stably maintained over multiple generations independently of PRG-1 [3] . Similarly , 22G-RNAs and silencing initiated by ingested dsRNA can be maintained in the germline for multiple generations without ingestion of dsRNA [3 , 13 , 16 , 17] . Transgenerational maintenance of silencing requires nuclear RNAi and chromatin factors . However , as these components are required as effectors of silencing by RNAi , it is unclear which of these factors play an active role in transmission of epigenetic information between generations . An important observation in defining the inherited signal behind epigenetic inheritance of RNAi-mediated silencing is that transgenerational silencing is associated with maintenance of high levels of 22G-RNAs , even several generations after exposure to the initial trigger [3 , 4] . Thus , understanding the mechanism of transgenerational RNAi requires elucidation of how 22G-RNAs are amplified across generations . Considering the large number of progeny generated each generation , this must reflect synthesis of new siRNAs . However , such amplification would suggest that secondary 22G-RNAs can drive synthesis of additional similar ( tertiary ) small RNAs . As yet there is no direct evidence for tertiary siRNAs biogenesis in the germline of C . elegans . Indeed , a recent study characterizing siRNAs generated in response to exogenous dsRNA suggested that secondary 22G-RNAs have limited ability to initiate a tertiary siRNA response under the conditions tested [18] . Thus there is a conflict between the observations of stable 22G-RNA populations and the lack of evidence for a mechanism that could generate them . Understanding this is crucial to understanding how small RNAs are involved in the maintenance of epigenetic states . In this study , we gain important insight into the resolution of this paradox by addressing the mechanism whereby 22G-RNAs are generated at piRNA target genes . We find that , in fact , a tertiary 22G-RNA population is found at piRNA target genes . Tertiary 22G-RNAs are formed downstream of secondary 22G-RNAs and are not directly dependent on initial target recognition . Crucially , this population is dependent on the activity of the nuclear RNAi pathway , in particular requiring the engagement of the secondary 22G-RNAs with the nuclear Argonaute HRDE-1 for their formation . Additionally , we find that these 22G-RNAs are necessary and sufficient to instigate a form of paramutation [19] whereby a stably silenced allele is generated that can in turn silence further alleles in trans independently of the initial trigger . Our work reveals a feed-forward siRNA amplification loop that guides epigenetic inheritance in animals .
We and others previously showed that the nuclear RNAi pathway is required for silencing of piRNA targets in C . elegans [3 , 4] . Utilizing a reporter transgene consisting of a GFP sequence upstream of a target site for the endogenous C . elegans piRNA 21UR-1 ( the piRNA sensor ) , we were able to detect secondary siRNAs generated near the piRNA target site in nrde-1 mutant animals [3] , suggesting nuclear RNAi acts downstream of secondary siRNA synthesis to silence piRNA targets . To investigate the role of the nuclear RNAi pathway more carefully , we carried out deep sequencing of small RNAs in wildtype and nuclear RNAi defective strains carrying the piRNA sensor . To our surprise , we found that although proximal 22G-RNAs localizing to within 200bp of the piRNA target site are still present , distal 22Gs mapping to the remainder of the GFP coding sequence ( as observed in wild type animals , Fig . 1A ) are lost in hrde-1 , nrde-1 and nrde-4 mutants . In contrast , prg-1 mutants lack both proximal and distal 22G-RNAs against the piRNA sensor mRNA , as do mutants lacking the mutator protein MUT-16 , which is required for synthesis of many 22G-RNAs ( Fig . 1A ) . The requirement for the nuclear RNAi machinery suggested that 22G-RNAs mapping to the piRNA sensor might bind to the nuclear Argonaute HRDE-1 . To test this we sequenced small RNAs associated with HRDE-1 using immunoprecipitation with a specific antibody ( Fig . 1B ) . Both proximal and distal 22G-RNAs were strongly enriched in the immunoprecipitated HRDE-1 complexes compared to a negative control immunoprecipitation ( Fig . 1B ) . Since only the distal 22G-RNAs were absent in hrde-1 mutants ( Fig . 1A ) , this implies that HRDE-1 binding to 22G-RNAs mapping close to the piRNA target site is required for formation of 22G-RNAs mapping across the transcript . The generation of 22G-RNAs requires RdRPs that utilise a targeted mRNA as a template . We therefore performed high-throughput sequencing of small RNAs from animals mutant in RdRPs carrying the piRNA sensor or an alternative mCherry piRNA sensor transgene and mapped 22G-RNA antisense reads to the transgene . As compared to wild type animals , single gene mutants of rrf-1 , rrf-2 , rrf-3 and ego-1 showed no reduction in either class of 22G-RNAs ( S1 Fig ) . However , the double mutant of rrf-1 and ego-1 showed a severe reduction of both proximal and distal 22G-RNAs mapping to the piRNA sensor , consistent with the observation that mutations in both genes are necessary to desilence a piRNA target [11] . In addition to RdRPs , the Dicer-related helicase DRH-3 was reported to be essential for the production of most 22G-RNAs . Initial high-throughput sequencing of a helicase-dead mutant of drh-3 suggested a function for DRH-3 in the 3′ to 5′ spreading of 22G-RNAs across endogenous genes [20] reminiscent of our observations in the nrde mutants . However , when using the same helicase mutant line to analyse small RNAs mapping antisense to the piRNA sensor , we observed loss of both proximal and distal 22G-RNAs , analogous to the rrf-1 ego-1 double mutant ( S1 Fig ) . Thus neither the helicase domain of DRH-3 nor one particular RdRP is required specifically for the generation of the distal 22G-RNAs . These observations were made using a transgene bearing a single piRNA target site in the 3′UTR . We therefore wanted to test whether proximal and distal 22G-RNAs could be detected at endogenous piRNA targets . We observed a similar pattern of 3′ to 5′ spreading of 22G-RNAs at the endogenous transcript Y48G1BM . 5 ( Fig . 2 ) . The presence of the distal 22G-RNAs depends on hrde-1 , nrde-1 and nrde-4 , while proximal 22G-RNA reads are also greatly reduced in prg-1 and mut-16 ( Fig . 2 ) . In contrast , nrde mutants do not show a reduction in 22G-RNAs at previously characterized endogenous piRNA targets . Since endogenous piRNA target sites could be distributed throughout the length of each target gene , we hypothesized that detection of proximal and distal 22G-RNAs would only be possible for specific piRNA target genes where there was a higher density of piRNA target sites at one or other end of the gene . Only these targets would show a clear reduction in 22G-RNAs in nrde mutants . To test this hypothesis we compared the distribution of 22G-RNAs along the length of all genomic transcripts by normalizing to transcript length . We then used k-means clustering to group all 22G-targeted transcripts into 7 clusters based on the distribution of 22Gs along the gene body in nuclear RNAi defective ( nrde ) and hrde-1 mutants ( see methods ) and compared the average abundance of 22G-RNAs as a function of transcript position between wild type and nrde-4 mutants for each cluster . Cluster 6 , containing 186 genes , showed strongly reduced 22G reads at the 5′ end of the transcripts compared to the 3′ end in nrde-4 and hrde-1 mutants ( Figs . 3A and S2 ) . Many individual genes from cluster 6 showed a pattern of nuclear RNAi dependent small RNA accumulation similar to that observed for the piRNA sensor ( such as Y48G1BM . 5 , Fig . 2 ) . This suggested that genes within this cluster might display distal 22G-RNAs similar to those seen in the piRNA sensor at their 5′ ends and proximal 22G-RNAs at their 3′ ends of the gene . This would imply that the sites directly targeted by piRNAs would be more common at the 3′ end within this cluster of genes . To test this we mapped sites predicted to be targeted by known piRNA sequences , allowing up to 3 mismatches [11] . Out of all 7 clusters , only Cluster 6 showed a statistically significant enrichment of piRNA target sites in the 3′ half of the gene compared to the 5′ half ( p = 0 . 03 , Fisher’s Exact Test; Fig . 3B ) . Furthermore , Cluster 6 showed strongly reduced 22G-RNA reads in prg-1 mutants relative to wild type , ( Fig . 3C; p<1e-16 , Wilcox signed rank test ) and showed enrichment for HRDE-1 immunoprecipitated 22G-RNAs ( Fig . 3D ) , consistent with recognition of these targets in vivo by both PRG-1 and HRDE-1 . Interestingly , cluster 1 showed modestly reduced 22G-RNAs at the 3′ end in nuclear RNAi mutants ( Fig . 3A ) , an enrichment of piRNA targets in the 5′ half of the gene ( Fig . 3B ) , and also had reduced 22G-RNAs in prg-1 mutants ( p<1e-12 , Wilcox signed rank test , Fig . 3C ) and was enriched for HRDE-1-immunoprecipitated 22G-RNAs ( Fig . 3D ) . This suggests the nuclear RNAi pathway may also mediate 22G-RNA spreading in a 5 to 3 direction along piRNA target transcripts . We conclude for at least a subset of endogenous piRNA targets 22G-RNAs distal to piRNA target sites are dependent on the nuclear RNAi pathway . This analysis supports the notion that two distinct classes of 22G-RNAs are made against many endogenous piRNA targets , and may play an important role in their regulation . The data above was strongly in support of a multi-stage model for the generation of 22G-RNAs against piRNA targets genome-wide . In such a model , proximal 22G-RNAs generated by the initial target recognition would in turn associate with HRDE-1 to trigger distal 22G-RNAs synthesis , in a manner dependent on the nuclear RNAi proteins NRDE-1 , NRDE-2 and NRDE-4 . This would classify 22G-RNAs that map across the entire transcript as “tertiary” siRNAs as opposed to secondary siRNAs generated by the initial target recognition event [18] . However , it is also possible that both classes are directly dependent on the initial target recognition event and occur in parallel . We therefore generated a reporter gene to distinguish between these two models . We took advantage of our previous observation that the piRNA-dependent 22G-RNAs present in animals carrying the piRNA sensor can trigger trans-silencing of a second GFP-encoding transgene that lacks a perfectly complementary piRNA target site [3] . This effect is mediated by siRNAs antisense to GFP , since a control cross to a similar mCherry-expressing transgene did not result in silencing ( S1 Table ) . We designed a transgenic operon that ubiquitously expresses both mCherry and GFP mRNAs from a single primary transcript , placing mCherry upstream of GFP within the operon ( Fig . 4A and 4B; hereafter referred to as the operon ) . Animals carrying a single-copy insertion of the operon transgene on chromosome I express both mCherry and GFP throughout development , in both somatic and germline tissues ( Fig . 4C ) . Expression of this transgene is stable and we did not observe any germline silencing in animals carrying the operon alone . Since the piRNA sensor gives rise to 22G-RNAs antisense to GFP we expected to observe silencing of GFP expression from the operon in animals carrying both the operon and the piRNA sensor . Further , if nuclear RNAi promotes spreading of siRNAs along the length of the target transcript , even in the absence of a perfectly complementary piRNA target site , we would expect expression of mCherry from the operon to be silenced , with concomitant synthesis of tertiary 22G-RNAs against to the mCherry coding sequence . Any such silencing and siRNA spreading must be triggered in the nucleus , since the mCherry and GFP mRNAs are exported to the cytoplasm separately ( see schematic Fig . 4A and 4B ) . Animals homozygous for the operon and the piRNA sensor ( a single-copy insertion on chromosome II ) , showed germline-specific silencing of GFP expression from the operon , confirming that the piRNA sensor-derived 22G-RNAs antisense to GFP are able to recognize the operon’s GFP-encoding sequence in trans . We also observed germline-specific silencing of mCherry expression ( Fig . 4C ) . Trans-silencing of the operon by 22G-RNAs derived from the piRNA sensor results in the synthesis of 22G-RNAs antisense to the mCherry coding sequence that are not normally found in animals carrying the operon transgene alone ( Fig . 4D ) . Since silencing of the operon transgene is not elicited by direct piRNA targeting , but by secondary 22G-RNAs derived from the piRNA sensor , these new 22G-RNAs antisense to mCherry must be tertiary siRNAs . The fact that spreading occurs across distinct transcripts within an operon indicates that tertiary siRNA synthesis is initiated in the nucleus , consistent with a role for the nuclear RNAi machinery . Consistent with the notion that nuclear 22G-RNAs elicit silencing , expression of both GFP and mCherry is restored in operon; piRNA sensor animals lacking hrde-1 ( Fig . 4C ) , mut-16 , nrde-1 , or nrde-4 ( S3A Fig ) . Along with the loss of trans-silencing , these animals also fail to generate tertiary 22G-RNAs antisense to the mCherry coding sequence ( Figs . 4D and S3B ) , confirming that nuclear RNAi is necessary to initiate synthesis of this population of tertiary siRNAs . It is important to note that although secondary 22G-RNAs and nuclear RNAi promote tertiary 22G-RNA biogenesis , these data do not indicate the site of ( presumably RdRP-dependent ) synthesis of new tertiary 22G-RNAs . These data demonstrate that secondary 22G-RNAs initiated by a piRNA target site can promote the synthesis of tertiary 22G-RNAs , and suggest secondary 22G-RNAs act in the nucleus and independently of the initial piRNA target site to promote tertiary siRNA biogenesis . These data therefore provide an explanation for the nuclear RNAi-dependent spreading of 22G-RNAs across piRNA target loci . The 22G-RNAs that silence piRNA targets , at least at loci with a single piRNA target site , can be divided into secondary and tertiary classes based on the following criteria: secondary 22G-RNAs are found proximal to the piRNA target site , and their biogenesis and stability does not require nuclear RNAi; tertiary siRNAs are found distal to the piRNA target site and are lost in nuclear RNAi-defective mutants . In the case of the piRNA sensor , the proximal and distal 22G-RNAs we identified are secondary and tertiary 22G-RNAs respectively . Nuclear RNAi mutants do not generate tertiary siRNAs and fail to silence the piRNA sensor , or to trans-silence the GFP-encoding operon , suggesting either that both secondary and tertiary siRNAs are required for efficient silencing or that tertiary siRNA generation itself is somehow coupled to the silencing process . Altogether , these data define a new class of tertiary siRNAs and suggests a new mode of small RNA amplification important for repression of piRNA targets . Both piRNAs and exogenous dsRNA can initiate multigenerational epigenetic silencing in the germline , which depends on nuclear RNAi , and is associated with maintenance of high levels of 22G-RNAs [3 , 4] . The tertiary 22G-RNAs we identified provide a mechanism for maintenance of high levels of 22G-RNAs over many generations as they could repeatedly re-engage the nuclear RNAi pathway to trigger synthesis of further HRDE-1 bound 22G-RNAs , thus generating a feed-forward amplification loop that can maintain robust levels of 22G-RNAs independent of the initial trigger . In support of this notion , after outcross of operon; piRNA sensor animals the operon remains silenced in the germline regardless of the presence of the piRNA sensor ( Fig . 5A ) . This silencing is highly stable; we did not observe any animals revert to germline expression of the operon over more than 12 continuous generations in multiple independent silenced lines . We sequenced small RNA populations in such stably silenced lines , and found that silencing is associated with 22G-RNAs antisense to the mCherry coding sequence ( Fig . 5B ) . Supporting the involvement of a nuclear RNAi-dependent 22G-RNA amplification loop , maintenance of this silencing requires the mutator proteins , nuclear RNAi factors , and HRDE-1 ( S4 Fig ) . In addition to triggering heritable operon silencing by the piRNA sensor , we also initiated heritable silencing by feeding the operon-expressing animals with bacteria expressing GFP dsRNA . RNA interference was effectively induced in the parental generation and led not only to complete silencing of the targeted GFP but also to silencing of mCherry . This transgene silencing was maintained over at least 4 generations after progeny of dsRNA-treated animals were removed from RNAi feeding plates . Concomitant with silencing of both the GFP and mCherry , we observed abundant 22G-RNAs antisense to the entire operon transcript in the dsRNA-treated P0 and their non-treated F2 generation indicating that exogenous dsRNA can also induce stable heritable gene silencing via a 22G-RNA amplification loop ( S5 Fig ) . Transitive RNAi or small RNA spreading as a consequence of dsRNA-triggered gene silencing seems to have a strong bias for a 3′ to 5′ direction with regard to the mRNA template [18 , 21] . However , our analysis of endogenous genes exhibiting nuclear RNAi-dependent spreading of small RNAs would support the notion that spreading could also occur in 5′ to 3′ direction ( Fig . 3 , cluster 1 ) . To test this directly using the operon transgene model , we made use of a second piRNA sensor transgene that encodes for a germline-expressed mCherry-Histone 2B fusion protein that also undergoes silencing via the endogenous piRNA 21UR-1 ( mCherry piRNA sensor , [11] ) . We repeated the trans-silencing experiment using the mCherry piRNA sensor to initiate silencing of the operon transgene ( S6A Fig ) . This led to the stable germline-specific silencing of both mCherry and GFP expressed from the operon . To investigate whether this silencing was accompanied by 22G-RNA reads mapping to the operon we sequenced small RNAs from silenced strains . We observed abundant small RNAs mapping to both mCherry and GFP portions of the operon , indicating that the 5′ to 3′ spreading of small RNAs can occur in this system ( S6B Fig ) . The stable silencing of the operon is reminiscent of paramutation , in which a silent allele initiates epigenetic silencing of an expressed allele [22 , 23] . The second allele is maintained in a silent state independently of the presence of the initiating allele , and moreover can induce silencing of further expressed alleles , thus itself has become paramutated . This ability to confer paramutation in trans is a critical property of paramutated alleles . We tested whether the stably silenced operon could act in this manner . We performed crosses between the silenced operon , and a number of stably germline-expressed transgenes ( Figs . 5C , 5E , S7 and S1 Table ) . Importantly , we never observed silencing in crosses between any two stably expressed transgenes ( Fig . 5E and S1 Table ) . However , crosses between the silenced operon and multiple transgenes bearing the mCherry coding sequence , including the operon itself , and a transgene that ubiquitously expresses an mCherry::H2A fusion under the spn-4 promoter ( hereafter mCherry::H2A ) ( Fig . 5C and E ) resulted in silencing of the expressed allele . After a cross to the silenced operon , silencing of the mCherry::H2A transgene was stably maintained in all descendants over >10 generations , regardless of the presence of the operon . This silencing was concomitant with the production of 22G-RNAs that map antisense to the mCherry sequence that are usually not observed in the constitutively expressing parental line ( Fig . 5D ) . Moreover the mCherry::H2A transgene , like the silenced operon , initiated stable silencing of other mCherry encoding transgenes . These observations are consistent with paramutation of the operon locus: the operon is stably silenced independent of the original trigger ( the piRNA sensor ) , and can confer such stable silencing in trans . Surprisingly , despite the robust activity of the silenced operon against mCherry expression , it was unable to induce silencing of GFP-expressing transgenes . Indeed , we were able to generate animals carrying both the stably silenced operon , and stably germline-expressed GFP::H2B fusions ( S7 Fig ) . This observation is puzzling since silencing of the operon was initiated by 22G-RNAs mapping to GFP . However , we note that animals carrying the paramutated operon have lower levels of 22G-RNAs antisense to GFP than animals carrying the piRNA sensor ( Figs . 1A and 5B ) . The lower levels of 22G-RNAs antisense to GFP derived from the silenced operon may be insufficient to initiate silencing . Differences in 22G-RNA levels may result from the sequence differences between the coding regions of the two transgenes leading to different abilities to act as a template for RdRP . It is also possible that other subtle differences in the organization of the transgene , such as promoter strength or choice of 3′ UTR , might lead to differences in their sensitivity to silencing . Taken together , our observations suggest that a silencing signal associated with the mCherry coding sequence promotes the paramutagenic activity of the stably silenced operon . Since we demonstrated that silencing of the operon locus requires nuclear RNAi-dependent synthesis of tertiary 22G-RNAs antisense to the mCherry coding sequence , this observation supports the notion that tertiary 22G-RNAs instigate silencing and paramutation . Moreover , the ability of the operon to trigger silencing in trans suggests that a diffusible signal is involved . Noting that HRDE-1/22G-RNA complexes are maternally contributed to the zygote , we speculated that inherited HRDE-1/22G-RNA complexes might be sufficient to initiate silencing ( and associated 22G-RNA amplification ) in progeny independent of the silenced operon locus itself . To test this possibility , we crossed hermaphrodites heterozygous for the silenced operon with males carrying a stably expressed mCherry::H2A transgene . We were able to distinguish cross progeny carrying or lacking the silenced operon based on somatic expression . We observed the same repression of mCherry::H2A expression in the germline of cross progeny regardless of inheritance of the silenced operon locus ( Fig . 5E ) . From progeny of animals that did not inherit the silenced operon locus we were able to isolate mCherry::H2A transgenic strains that maintained stable multigenerational silencing in the germline ( from F1 progeny shown in Fig . 5E , last panel on right ) . Thus , transmission of silencing does not require inheritance of the silenced locus , implying that altered chromatin at the silenced gene is not pivotal to transmit silencing across generations . Instead , a maternal copy of the silenced operon confers multigenerational silencing independent of its own transmission . Thus , our data suggest that inheritance of maternal HRDE-1/22G-RNA complexes is sufficient to confer multigenerational silencing and paramutation . Taken together , these data support a model in which nuclear RNAi-dependent feed forward amplification of tertiary 22G-RNAs is necessary and sufficient for paramutation in animals .
Gene silencing by small RNAs in C . elegans is characterised as a two-stage process , by which the primary siRNA:Ago complex recruits RdRPs to a target mRNA that in turn is used as a template for the production of unprimed 22G-RNAs [24 , 25] . However , so far there has been no evidence for a multi-stage mechanism to generate 22G-RNAs spreading across a target gene . Here we observed that a subset of 22G-RNAs , which map to regions distal to piRNA target sites on both transgenes and certain endogenous genes , specifically require the nuclear RNAi pathway including the nuclear Argonaute HRDE-1 . As these 22G-RNAs are produced only downstream of 22G-RNAs that map closer to piRNA target sites , these can be described as tertiary 22G-siRNAs . Although we cannot formally exclude the possibility that the nuclear RNAi pathway is required only for the stability of tertiary 22G-RNAs , the fact that HRDE-1 binds equally to both classes whilst only being required for the presence of tertiary 22G-RNAs ( Fig . 1 ) suggests that the nuclear RNAi pathway is required for the generation of tertiary 22G-RNAs . In the first stage of this process , secondary 22G-RNAs generated as a direct result of piRNA target recognition bind to the Argonaute HRDE-1 and are transported to the nucleus . Target recognition , in conjunction with the activities of the NRDE proteins NRDE-1 , NRDE-2 and NRDE-4 , subsequently results in the generation of further , tertiary , 22G-RNAs ( Fig . 6 ) . In contrast to the initial 22G-RNA population , tertiary 22G-RNAs are not restricted to the direct vicinity of the piRNA target site , but instead can spread throughout the transcript . Further experiments will be required to clarify the mode of tertiary 22G-RNA biogenesis . In particular , it is at present unclear how the nuclear RNAi pathway recruits RNA dependent RNA polymerase to the target transcript . One simple possibility is that HRDE-1 directly recruits the RdRP to engage the target transcript . However , we have no evidence for a direct interaction between HRDE-1 and any RdRPs , and the RdRPs previously implicated in 22G-RNA biogenesis localise to the cytoplasm [12] . Moreover , we detected almost no 22G-RNAs against the introns of the transgenes we used in our study ( S8 Fig ) suggesting that RdRPs act on spliced transcripts , in favour of a cytoplasmic location for tertiary 22G-RNA synthesis . This would require some form of indirect communication between nuclear RNAi factors such as HRDE-1 and the cytoplasmic 22G-RNA biogenesis machinery . The exact mechanism of this communication will be an exciting question for further research . The ability of tertiary 22G-RNAs to engage the Argonaute HRDE-1 and to instigate the generation of further 22G-RNAs was implied by the detection of abundant 22G-RNAs against germline-expressed transgenes silenced by piRNA targeting that remained in the absence of the initial piRNA itself [3 , 4] . However , seemingly in contradiction to these observations , in RNAi feeding experiments using dsRNA matching to ubiquitously expressed endogenous genes , careful analysis of deep sequencing data did not give evidence for abundant tertiary siRNAs [18] . Our data enables us to offer a resolution to this paradox . Here we show that the activity of nuclear RNAi in the germline enables secondary 22G-RNAs , produced as a direct result of piRNA target recognition , to generate tertiary 22G-RNAs . Thus we suggest that the very low abundance of tertiary siRNAs in the RNAi feeding experiments described by Pak et al . ( 2012 ) may reflect a lack of significant germline nuclear RNAi activity . This would be consistent with the fact that germline nuclear RNAi is not generally required for the efficacy of RNAi by feeding whilst the trigger is present [13] , and with the fact that RNAi targeting genes expressed in the soma is only rarely inherited for more than one generation [26] . It seems that , in addition to piRNA targeting , a subset of exogenous RNAi triggers are able to engage the germline nuclear RNAi pathway . Further work will be required to decipher the rules that govern which triggers end up in the nuclear RNAi machinery . This may relate to how some endogenous genes may be protected from targeting by the piRNA pathway as discussed further in the next section . As tertiary siRNAs can instigate the formation of further 22G-RNAs they could lead to potentially unlimited loops of silencing . Indeed we observed such effects in demonstrating paramutation whereby a silenced transgene can convert an expressed transgene into a silent one that in turn has the ability to silence further transgenes . A serious risk is that such an amplification system could silence genes erroneously through the amplification of off-target effects [18] . One way to reduce the likelihood of this process occurring would be the protection mechanism proposed to prevent most germline genes from becoming targets of the piRNA pathway [27 , 28] . The protection proposed involves the Argonaute CSR-1 , which binds to 22G-RNAs antisense to nearly all germline expressed genes , but rather than triggering silencing , prevents targets from being shut down , possibly by antagonizing HRDE-1 . In the absence of CSR-1 activity , HRDE-1 activity dominates leading to loss of expression [27 , 28] . Thus the formation of tertiary 22G-RNAs would only be able to occur at genes that are not targets of CSR-1 , such as foreign transgenes and certain endogenous genes including those we identify as being subject to NRDE-dependent siRNA spreading along the transcript . Similarly , RNAi-induced silencing of endogenous CSR-1 targets would be limited to short-term effects and would not engage HRDE-1 and thus would not result in tertiary 22G-RNA synthesis . Another layer of protection against unchecked silencing once it is established could be through restrictions in successive 22G-RNA production cycles . Given the many endogenous loci targeted by HRDE-1-associated 22G-RNAs , this could likely occur due to competition for a limited pool of factors required for 22G-RNA biogenesis and/or stability such as the RdRPs or Argonaute proteins eventually preventing their uncontrolled amplification . Given the fact that many C . elegans genes are protected from stable silencing , it is possible that the multigenerational silencing mediated by HRDE-1/22G-RNAs may only apply to transgenes such as the ones that we use in this study . However , here we describe a set of endogenous genes that behave similarly to the piRNA sensor such that 22G-RNAs mapping across the transcript are reduced specifically in mutants lacking the nuclear RNAi pathway whilst those in the vicinity of predicted target sites are maintained . Notably , the genes that fit into this category may represent only a fraction of the genes that are dependent on HRDE-1 for silencing , because 22G-RNAs mapping to HRDE-1 targets would not be expected to be affected if the piRNA target sites map along the entire gene rather than in one place . Recently , Ni et al . reported the identification of a set of HRDE-1 target genes that were subject to transcriptional silencing including the accumulation of repressive histone modifications [29] . In accordance with our prediction that 22G-RNAs may not be affected despite dependence on HRDE-1 for silencing , a subset of the genes identified by Ni et al . , showed little reduction in 22G-RNAs in hrde-1 mutants [29] . Thus we suggest that some endogenous genes may be subject to transgenerational silencing in addition to transgenes . Although changes in gene expression that could result from HRDE-1-dependent silencing of endogenous genes may be limited to the germline , this could be linked by systemic signalling to effects in the soma [30 , 31] . As such , it could provide a mechanism for the transgenerational inheritance of developmental or physiological traits in response to environmental triggers , for example starvation , elevated temperature , or exposure to pathogens . In the harsh world of rotting apples inhabited by C . elegans , many environmental changes occur on a slower timescale than C . elegans’ fast reproductive cycle , so ‘Lamarkian’ epigenetic inheritance could be advantageous .
We grew C . elegans under standard conditions at 20°C unless stated otherwise . The wild type strain was var . Bristol N2 [32] . The food source used was E . coli strain HB101 ( Caenorhabditis Genetics Center , University of Minnesota , Twin Cities , MN , USA ) . We used bleaching followed by starvation-induced L1 arrest to generate synchronized cultures . Detailed information about genetic crosses and a list of all strains generated and used in this study are in the Supporting Information ( S1 Text ) . Detailed information is provided in the Supporting Information ( S1 Text ) . RNAi experiments for transgenerational inheritance were essentially performed as described [3] . Briefly , three L4 larvae were plated onto RNAi plates seeded with either empty vector ( pL4440 ) or GFP dsRNA expressing HT115 ( DE3 ) bacteria and allowed to produce progeny . Efficiency of RNAi in this progeny was monitored by GFP fluorescence microscopy . Adult animals were removed after 5 days of initiating RNAi and transferred to non-RNAi plates . Progeny of these animals were sequentially transferred for 3 more generations and monitored for GFP expression . For high-throughput small RNA sequencing , ca . 50 animals from the P0 ( treated ) or the F2 ( non-treated ) generations were picked and subjected to RNA extraction . We carried out differential interference contrast ( DIC ) and fluorescence imaging by standard methods [33] using a ZEISS AX10/Imager . A1 upright microscope with 20x objective magnification . Images were taken using an ORCA-ER Digital Camera ( Hamamatsu ) and processed using Openlab 5 image software ( Improvision ) and Fiji/ImageJ ( version 1 . 48d ) . Synchronised animals were grown in liquid culture under standard conditions using HB101 as food source at 20°C and collected as young adults to gravid animals . Bacteria were washed off by several washes with M9 . Animals were washed and transferred into lysis buffer ( 20 mM Tris/Cl , ph 8 . 0 , 140 mM KCl , 1 . 8 mM MgCl2 , 0 . 5% NP-40 , 1 mM DTT , 0 . 1 mM PMSF , Protease Inhibitor Cocktail ) . A 1:1 slurry of animals in lysis buffer was snap frozen in liquid nitrogen , thawed and homogenized using 0 . 7mm diameter Zirconia beads ( BioSpec products ) in a Precellys 24 homogenizer ( Bertin Technologies ) at 6500 rpm for 2x 20 seconds . Cell debris was removed by centrifugation and 20 mg of cleared extracts was subjected to HRDE-1 immunoprecipitation using 2 . 5 μg anti-HRDE-1 polyclonal antibodies [3] . Antibody-antigen complexes were recovered using Dynabeads Protein A ( Life Technologies ) . 5% of immune-precipitated material was subjected to SDS-PAGE followed by Western Blotting with anti-HRDE-1 and anti-PRG-1 antibodies [34] . The remaining antibody-antigen complexes were eluted from the Dynabeads by adding 1ml TRIsure reagent ( Bioline ) to extract RNA . After snap-freezing in liquid Nitrogen , samples were processed according to the manufacturer’s protocol . For total RNA isolation we harvested synchronised young adult and gravid animals from plates by washing with M9 . Alternatively , for animals bearing mutations giving rise to sterile homozygous adults , we picked ca . 50 balancer GFP-negative homozygous mutant L4 to young adult animals from a balanced heterozygous population . The same number of age-matched control animals was used alongside for RNA preparations and small RNA library preparations . We pelleted and dissolved animals in 10 volumes of TRIsure reagent ( Bioline ) . After snap-freezing we homogenized animals by five freeze-thaw cycles in liquid nitrogen . We extracted total RNA according to the manufacturer’s protocol . cDNA libraries were prepared by treating 1–5 μg total RNA or RNA extracted after HRDE-1 immunoprecipitation with 20 Units RNA 5′ polyphosphatase ( Epicentre ) in a total volume of 20 μl . De-phophorylated RNA was purified by phenol-chloroform extraction and ethanol precipitation according to standard protocols . Subsequent library preparations were performed with the TruSeq Small RNA library kit ( Illumina ) following the manufacturer’s instructions with exception that 15 cycles of PCR amplification were used . We size-selected cDNA libraries using 6% TBE PAGE gels ( Life Technologies ) and ethidium bromide staining . Desired sizes of cDNA bands were cut from the gel ( between 147 and 157 nt ) , the gel matrix broken by centrifugation through gel breaker tubes ( IST Engineering Inc . ) , and cDNA eluted with 400 μl of 0 . 3M Na-Chloride . Further purification of cDNA was by centrifugation through Spin-X 0 . 22μm cellulose acetate filter columns ( Costar ) followed by ethanol precipitation . Libraries were sequenced on a MiSeq Benchtop Sequencer or a HiSeq 2500 Sequencer ( Illumina ) . Processing of small RNA sequencing data to obtain alignments to transcripts was essentially as described [35 , 36] . For analysis of the genome-wide distribution of small RNAs along transcripts , antisense 22Gs were counted in 10 bins evenly spaced along each spliced transcript corresponding to protein-coding genes as annotated in Wormbase ( WS190 ) . These data in nrde-4 mutants were used as the input for k-means clustering . The algorithm was repeated using 2–8 clusters; at 8 clusters the algorithm failed to converge , thus 7 clusters were used for the final analysis . The genes identified in each cluster were then selected in nrde-4 and wild type and the average number of reads in each bin in nrde-4 , normalized to total library size , was divided by the average number in wild type , normalized to total library size to provide the relative ratio of nrde-4 to wild type at each bin . The logarithm of this was plotted in Fig . 3 . The total number of antisense 22Gs across the entire transcript , normalized for total library size was also obtained for each gene . These values were used to plot Fig . 3C and 3D . piRNA target prediction was done by selecting for piRNAs that aligned with up to 3 mismatches antisense to transcripts using Bowtie for alignment as described previously [11] . All data analysis was performed in the R programming environment . | Transgenerational epigenetic gene silencing has been shown to be important for organisms to react directly to their environment without the need to acquire genetic mutations . The inheritance of acquired traits via the gametes can prove advantageous in fast reproducing organisms . In Caenorhabditis elegans , a free-living nematode , multigenerational epigenetic inheritance can be induced by exogenous ( experimentally provided ) and endogenous cues that trigger small RNA-dependent gene silencing in the germline of these animals . PIWI interacting small RNAs ( piRNAs ) are required for the initiation of stable silencing of invading genomic elements in the germline such as transposons . Gene silencing established by piRNAs can subsequently be maintained over multiple generations without the original trigger . In C . elegans , this stable maintenance of silencing requires an additional class of small interfering RNAs ( siRNAs ) that must be amplified in each generation in order to maintain multigenerational silencing . Here we show that these siRNAs fall into two distinct classes , which we call secondary and tertiary siRNAs . We find that the production of tertiary siRNAs is part of a nuclear amplification pathway associated with the stable heritable silencing of an allele , a form of paramutation . This amplification pathway therefore promotes germline integrity and possibly the inheritance of acquired physiological traits . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2015 | Tertiary siRNAs Mediate Paramutation in C. elegans |
In this study the ‘Malaria Box’ chemical library comprising 400 compounds with antiplasmodial activity was screened for compounds that perturb the internal pH of the malaria parasite , Plasmodium falciparum . Fifteen compounds induced an acidification of the parasite cytosol . Two of these did so by inhibiting the parasite’s formate nitrite transporter ( PfFNT ) , which mediates the H+-coupled efflux from the parasite of lactate generated by glycolysis . Both compounds were shown to inhibit lactate transport across the parasite plasma membrane , and the transport of lactate by PfFNT expressed in Xenopus laevis oocytes . PfFNT inhibition caused accumulation of lactate in parasitised erythrocytes , and swelling of both the parasite and parasitised erythrocyte . Long-term exposure of parasites to one of the inhibitors gave rise to resistant parasites with a mutant form of PfFNT that showed reduced inhibitor sensitivity . This study provides the first evidence that PfFNT is a druggable antimalarial target .
The most virulent malaria parasite , Plasmodium falciparum , was responsible for the majority of the 214 million malaria cases and 438 , 000 malaria-attributable deaths estimated to have occurred in 2015 [1] . The parasite has developed resistance to most of the drugs deployed against it [2] , and it is imperative that new drugs be developed to protect or replace existing therapies . Significant progress has been made in recent years in developing new antimalarials [3]; however , many of the compounds under development are structurally related to previous or current antimalarial drugs and there is an urgent need to identify novel lead-drug compounds that act on hitherto unexploited parasite targets and create improved options for resistance-deterring combination therapies . Such therapies should , ideally , contain at least two drugs that act on separate targets [4] . Recent large-scale whole cell phenotypic screens have uncovered tens of thousands of novel inhibitors of the in vitro growth of asexual P . falciparum parasites in human erythrocytes [5–7] . In a bid to further research into the novel antiplasmodial chemotypes , the Medicines for Malaria Venture ( MMV ) compiled the open access ‘Malaria Box’ , a collection of 400 structurally-diverse compounds selected from the drug screen hits for which the mechanisms of action were not known [8] . Determining the mechanisms of action of these antiplasmodial compounds has the potential to uncover aspects of P . falciparum biology that can be exploited as drug targets . The mechanism by which P . falciparum generates ATP presents potential vulnerabilities . In particular , the ( disease-causing ) asexual intraerythrocytic stages of P . falciparum rely primarily on glycolysis for energy metabolism [9] . Human erythrocytes infected with trophozoite-stage parasites devour glucose up to 100 times faster than uninfected erythrocytes [10] , and generate large quantities of lactate , which is exported via a lactate:H+ symport mechanism [11] . Both glucose uptake and lactate export are likely to be essential for maintaining the energy requirements , intracellular pH and osmotic stability of these parasite stages . The P . falciparum hexose transporter ( PfHT ) mediates the uptake of glucose into the parasite and has been of interest as a drug target for some time [12 , 13] . Recently , a compound from the Malaria Box was found to inhibit PfHT potently ( with a 50% inhibitory concentration ( IC50 ) of ~ 50 nM ) while also displaying a high degree of selectivity for PfHT over the human glucose transporter GLUT1 [14] . A transport protein that mediates the efflux of lactate from the intraerythrocytic malaria parasite has recently been identified and characterised [15–17] . The protein belongs to the microbial formate nitrite transporter ( FNT ) family , localises to the parasite plasma membrane ( as well as to the membrane bounding the parasite’s internal digestive vacuole ) , and transports lactate , as well as a variety of other monocarboxylates , in a pH-dependent manner that is consistent with H+-coupled transport [16 , 17] . The FNT family is structurally unrelated to the monocarboxylate transporters that export lactate from human cells [16 , 17] . In this study we have screened the Malaria Box for compounds that alter the intracellular pH of asexual trophozoite-stage parasites . Fifteen of the 400 compounds were found to acidify the parasite cytosol . Of these , two compounds were found to exert their effects on the parasite’s cytosolic pH by targeting the H+-coupled efflux of lactate via PfFNT .
In a previous study we showed that 28 compounds from the Malaria Box , when added to isolated asexual P . falciparum parasites , give rise to a cytosolic alkalinisation , together with an increase in the cytosolic [Na+] [18] . The data are consistent with the compounds inhibiting the putative Na+/H+ ATPase PfATP4 [19] . Here , we screened the remaining 372 Malaria Box compounds for their effects , at 1 μM , on parasite cytosolic pH . The screen entailed adding a 1 μL aliquot of a 1 mM stock of each compound to a 1 mL suspension of trophozoite-stage P . falciparum parasites that had been isolated from their host erythrocytes by brief exposure to saponin and loaded with the pH-sensitive fluorescent dye BCECF . The fluorescence of the parasite suspension was monitored continuously . If , on addition of a compound , no change in fluorescence was observed after 2 min , another compound was added to the same cell suspension . A maximum of 9 compounds were added successively to a single batch of cells before concanamycin A ( 100 nM ) , an inhibitor of the parasite’s H+-extruding V-type ATPase that has been shown previously to acidify the cytosol in P . falciparum [20] , was added as a positive control to confirm that a pH change was still detectable in those cells ( Fig 1A ) . Fifteen of the 372 Malaria Box compounds tested were found to give rise to an immediate-onset gradual decrease in fluorescence ratio ( i . e . a decrease in the ratio of the fluorescence measured using two excitation wavelengths: 495 nm [pH-sensitive numerator] and 440 nm [pH-insensitive denominator] ) . In each case the decrease in fluorescence ratio resulted from a decrease in the pH-sensitive numerator , consistent with the cells having undergone cytosolic acidification ( Fig 1B ) . The 15 compounds that induced an acidification in the initial screen ( and for which structures are shown in S1 Table ) were all re-tested in at least two further pH experiments and their effects on the cytosolic pH of the parasite confirmed . The 15 compounds were also tested for their effects ( at 1 μM and 5 μM ) on the pH of the parasite’s digestive vacuole . This organelle is acidic , with an estimated pH value of approximately 5 [21–23] . Seven of the 15 compounds dissipated the pH gradient across the digestive vacuole membrane ( causing its alkalinisation; S1 Table; S1 Fig ) . Of the 372 Malaria Box compounds tested for their effects on cytosolic pH , 23 gave rise to effects on the fluorescence ratio that were considered likely to be ‘optical effects’ rather than genuine pH changes . Such effects arise from either an intrinsic fluorescence of the compound or an interaction between the compound and the fluorescent dye . In each case the change of the fluorescence ratio was abrupt rather than gradual ( see example in Fig 1C ) . Furthermore , in most ( 20 out of 23 ) cases the change in fluorescence ratio resulted in full or in part from a change in the nominally pH-insensitive fluorescence emanating from excitation at 440 nm ( the denominator ) . After an optical effect was encountered , concanamycin A ( 100 nM ) was added to the cells to determine whether a pH change was still detectable ( which , in each case , it was , ruling out the possibility that the compound that caused the optical effect had also dissipated the pH gradient across the plasma membrane ) , and the cells were replaced with a new batch of cells before screening further Malaria Box compounds . One mechanism by which a compound might induce a cytosolic acidification is through inhibition of one ( or more ) of the plasma membrane proteins that mediate the efflux of H+ from the parasite . Lactate , produced in large quantities by the parasite as the end-product of glycolysis , is excreted via a coupled lactate:H+ transporter [11 , 16 , 17] , raising the possibility that this protein might be a candidate target for some of the compounds identified as hits in the initial screen . Compounds that inhibit the lactate:H+ transporter might be expected to: ( i ) have no effect on cytosolic pH in parasites suspended in a glucose-free saline , as under this condition glycolysis is not operating , and there is therefore no production of lactate; ( ii ) prevent the lactate:H+ transporter-mediated cytosolic acidification that results from the exposure of a parasite to a large inward lactate gradient . Although the physiological role of the transporter is to mediate the net export of lactate , down the normally-outward lactate concentration gradient , the transporter is bidirectional and , under conditions of an imposed inward lactate concentration gradient ( as results from the addition of a high concentration of lactate to the extracellular medium ) it mediates a net influx of lactate:H+ , resulting in a cytosolic acidification [11] . When tested at concentrations of 2 . 5–5 μM , only two of the 15 pH-lowering hits , MMV007839 and MMV000972 ( which are structurally similar to one another; S1 Table ) demonstrated both features that would be expected from a complete inhibition of lactate:H+ transport ( Fig 2 ) . The addition of MMV007839 or MMV000972 to isolated parasites suspended in a glucose-containing medium ( in which glycolysis was active ) resulted in cytosolic acidification ( Fig 2B and 2C ) . The exposure of parasites to the solvent alone ( 0 . 1% v/v DMSO ) had no effect on pH ( Fig 2A ) . When isolated parasites are suspended in glucose-free medium , glycolysis ceases , the intracellular ATP concentration decreases to zero , and the parasite is unable to maintain the activity of the plasma membrane V-type H+-ATPase which serves to generate and maintain an inward H+ electrochemical gradient [24 , 25] . As a result , the resting pH decreases to 7 . 0–7 . 1 ( i . e . close to the extracellular pH ) compared with 7 . 2–7 . 3 in glucose-replete cells . On addition of MMV007839 or MMV000972 to isolated parasites suspended in glucose-free medium ( in which lactate production is eliminated ) there was no change in the resting pH ( Fig 2E and 2F ) , consistent with the acidification seen in glucose-replete cells being due to the inhibition of lactate efflux . On restoration of glucose to parasites suspended in glucose-free medium ( in the absence of inhibitors ) pH increased as ATP levels , and hence the activity of the H+-extruding V-type H+-ATPase , were restored ( Fig 2D ) . When the same manoeuvre was applied to parasites exposed to either MMV007839 or MMV000972 , there was an initial acidification ( consistent with lactate production having been restored and with the compounds inhibiting the efflux of the newly-generated lactate:H+; Fig 2E and 2F ) . However , after ~ 2 min the cells underwent a gradual alkalinisation , restoring the cytosolic pH to approximately 7 . 1 ( Fig 2E and 2F ) . The basis for this recovery is not understood; no such recovery was observed when the compounds were added to parasites maintained continuously in the presence of glucose ( Fig 2B and 2C ) . MMV007839 and MMV000972 were then tested for their ability to inhibit the lactate:H+ transporter-mediated acidification of the parasite cytosol that normally occurs when lactate is added to the extracellular medium . This was tested at 4°C ( Fig 2G–2I ) to reduce the ability of the H+-extruding V-type H+-ATPase to counteract the lactate-mediated pH change [11] . At 4°C the acidification of the parasite cytosol induced by the addition of MMV007839 and MMV000972 was less pronounced than that seen at 37°C ( cf . Fig 2B and 2C and Fig 2H and 2I ) , reflecting the reduced rate of glycolysis and hence reduced rate of lactate production at the lower temperature . The addition of 10 mM L-lactate to parasites ( following a pre-treatment with solvent ( 0 . 25% v/v DMSO ) , which had no effect on cytosolic pH ) induced an abrupt acidification of the parasite cytosol ( Fig 2G ) , consistent with lactate entering the parasite rapidly in symport with H+ . On addition of 10 mM L-lactate to parasites that had been pre-treated with MMV007839 or MMV000972 , there was no such abrupt acidification ( Fig 2H and 2I ) . Rather , the slow acidification mediated by the compounds continued at the same rate as that seen prior to the addition of L-lactate . For MMV007839 the rates of acidification before and after the addition of lactate were 0 . 011 ± 0 . 003 pH unit/min ( mean ± SEM ) and 0 . 011 ± 0 . 004 pH unit/min ( n = 3; P = 0 . 9; paired t-test ) . For MMV000972 the rates were 0 . 012 ± 0 . 003 pH unit/min and 0 . 015 ± 0 . 006 pH unit/min ( n = 3; P = 0 . 5 ) . These data are consistent with the two MMV compounds inhibiting the lactate:H+ transporter and , thereby , the influx of lactate:H+ into the parasite . The pH data obtained with the isolated parasites provide indirect evidence that MMV007839 and MMV000972 inhibit lactate:H+ symport across the parasite plasma membrane but do not reveal the molecular identity of the target . To gain information on the possible molecular target of the compounds , we cultured a recently-cloned P . falciparum Dd2 parasite line in the presence of increasing concentrations of MMV007839 until parasites showing resistance to the growth-inhibiting effect of this compound emerged . Two independent cultures ( referred to here as ‘MMV007839-selected cultures A and B’ ) were initially maintained in the presence of 100 nM MMV007839 , and the concentration subsequently increased incrementally over the course of six weeks . The parasites from both cultures were tested , along with the Dd2 parental line , for their sensitivity to growth inhibition by MMV007839 and MMV000972 . Dd2 parental parasites had an IC50 value for MMV007839 of 0 . 14 ± 0 . 02 μM and an IC50 value for MMV000972 of 1 . 8 ± 0 . 2 μM . Parasites from both MMV007839-selected cultures were highly resistant to both compounds , with IC50 values > 280-fold higher than that of the Dd2 parent for MMV007839 , and > 30-fold higher than that of the Dd2 parent for MMV000972 ( P < 0 . 001; paired t-tests; Table 1 ) . The responses of parasites in the MMV007839-selected cultures to chloroquine and artemisinin were not significantly different from those of the Dd2 parental parasites ( P > 0 . 1; Table 1 ) . Genomic DNA was extracted from both of the MMV007839-selected cultures as well as from the parental Dd2 culture . The entire PfFNT gene was then amplified by PCR and the coding regions sequenced . Parasites from each of the MMV007839-selected cultures were found to have a mutation that was not present in the Dd2 parental parasites or in the 3D7 reference sequence . In both MMV007839-selected cultures ( A and B ) , the mutation in the coding sequence was G319A , which introduces a Gly107Ser mutation into the PfFNT protein . The finding of a mutation in the lactate:H+ transporter PfFNT in the MMV007839-selected cultures is consistent with PfFNT inhibition being the mechanism of action of MMV007839 and MMV000972 . The parental and MMV007839-selected Dd2 parasites also had two polymorphisms in the PfFNT coding sequence relative to the 3D7 strain ( PlasmoDB ID PF3D7_0316600 ) : a C475G change , which codes for a His159Asp mutation , and a A756G change ( synonymous ) . The effects of MMV007839 and MMV000972 on lactate transport across the parasite plasma membrane were investigated by measuring the uptake of L-[14C]lactate by saponin-isolated trophozoite-stage parasites in the presence and absence of the compounds . As noted above , although the physiological role of the transporter is to facilitate the net export of lactate from the parasite , the transporter is bidirectional , and measuring the unidirectional flux of L-[14C]lactate into the parasite provides a convenient means of monitoring the activity of the transporter . It has been shown previously that the uptake of external L-[14C]lactate by isolated parasites is enhanced at lower pH values , consistent with the transporter being H+-coupled [11] . The L-[14C]lactate uptake experiments were therefore performed on isolated parasites suspended in an acidic ( pH 6 . 1 ) medium , at 4°C ( to slow the transport process ) . The level of intracellular L-[14C]lactate accumulation was determined 20 s after the addition of the radiolabel to parasites . Under these conditions ( i . e . pH 6 . 1 , 4°C ) the 20 s incubation falls within the initial linear phase of uptake of the radiolabel and the measured uptake therefore reflects the lactate influx rate ( our own preliminary experiments and [11] ) . In isolated 3D7 parasites to which had been added DMSO ( 0 . 4% v/v , as a solvent control ) the L-[14C]lactate ‘distribution ratio’ ( i . e . the intracellular L-[14C]lactate concentration divided by the extracellular L-[14C]lactate concentration ) at 20 s reached a value of 6 . 7 ± 0 . 4 ( mean ± SEM; n = 4; S2 Fig ) . As has been reported previously [16] , the broad-specificity anion transport inhibitor NPPB ( 50 μM ) slowed L-[14C]lactate influx , reducing the distribution ratio at 20 s to 0 . 4 ± 0 . 1 ( P < 0 . 001; one-way ANOVA with post hoc Tukey test; S2 Fig ) . MMV007839 ( 0 . 5 μM ) inhibited L-[14C]lactate influx to a similar degree ( P < 0 . 001; S2 Fig ) , with 2 μM MMV007839 , as well as 0 . 5 μM and 2 μM MMV000972 , causing complete inhibition of L-[14C]lactate uptake ( P < 0 . 001; S2 Fig ) . A comparison of the uptake of L-[14C]lactate into the PfFNTGly107Ser mutant parasites ( from MMV007839-selected culture B ) and their parental parasites revealed that transport was slowed significantly in the mutant strains ( P < 0 . 001; unpaired t-test; Fig 3A ) . In the PfFNTGly107Ser mutant parasites the L-[14C]lactate distribution ratio , as measured at 20 s , reached a value of 2 . 1 ± 0 . 2 ( mean ± SEM; n = 7 ) , compared to a value of 5 . 3 ± 0 . 5 ( n = 9 ) in the parental parasites . This finding suggests that the Gly107Ser mutation in PfFNT causes some impairment of its function . NPPB ( 100 μM ) reduced the rate of L-[14C]lactate influx into both PfFNTGly107Ser mutant and parental parasites ( Fig 3A; P < 0 . 001; unpaired t-tests ) . The concentration dependence of the inhibition of L-lactate influx by the two MMV compounds in the PfFNTGly107Ser mutant parasites ( MMV007839-selected culture B ) was compared to that in the parental line . Both compounds were potent inhibitors of L-[14C]lactate uptake by Dd2 ( parental ) parasites ( Fig 3B and 3C ) . The IC50 values for MMV007839 and MMV000972 in these experiments were 158 ± 42 nM ( mean ± SEM; n = 7 ) and 49 ± 14 nM ( n = 3 ) , respectively . In the PfFNTGly107Ser mutant parasites the efficacy of MMV007839 and MMV000972 at inhibiting L-[14C]lactate transport was significantly reduced ( Fig 3B and 3C; P ≤ 0 . 001; unpaired t-tests ) with IC50 values of 16 . 3 ± 4 . 4 μM ( n = 3 ) and 15 . 2 ± 1 . 9 μM ( n = 4 ) , respectively . The finding of a reduced rate of L-[14C]lactate transport across the plasma membrane of PfFNTGly107Ser mutant parasites compared to parental parasites ( Fig 3A ) raised the question of whether the Gly107Ser mutation in PfFNT might be associated with a reduction in parasite fitness . To investigate this , we compared the growth rates of two PfFNTGly107Ser mutant parasite clones ( generated from MMV007839-selected culture B by limiting dilution ) with that of the parental Dd2 clone . We performed competition experiments in which mutant and parental parasites were mixed in an approximately 1:1 ratio , and their relative proportions monitored over time . These experiments revealed that the proportion of mutant and parental parasites remained approximately constant over the course of three weeks ( S3 Fig ) . Thus , the Gly107Ser mutation in PfFNT was not associated with a decrease in the growth rate of asexual blood-stage parasites , suggesting that the parasite can withstand a substantial reduction in the rate of PfFNT-mediated lactate transport before its viability is compromised . To determine directly whether PfFNT is inhibited by MMV007839 and MMV000972 , and if so whether the Gly107Ser mutation reduces the sensitivity of PfFNT to inhibition by the compounds , we expressed native PfFNT and PfFNTGly107Ser in Xenopus laevis oocytes . Consistent with our observations in parasites ( Fig 3A ) , which revealed that the rate of L-[14C]lactate transport into PfFNTGly107Ser mutant parasites was only 39 ± 5% ( mean ± SEM; n = 7 ) of that observed in the parental parasites , Xenopus oocytes expressing the mutant PfFNTGly107Ser protein were found to be significantly impaired in their ability to transport L-[14C]lactate relative to oocytes expressing native PfFNT ( Fig 4 ) . In the absence of MMV007839 or MMV000972 , the PfFNT-mediated component of L-[14C]lactate transport into oocytes expressing PfFNTGly107Ser was only 52 ± 7% ( mean ± SEM; n = 4 ) of that measured in oocytes expressing native PfFNT ( P < 0 . 001; one-way ANOVA with post hoc Tukey test ) . MMV007839 and MMV000972 inhibited L-[14C]lactate uptake via both the native PfFNT and PfFNTGly107Ser transporters ( Fig 4 , S4 Fig ) , whilst causing no changes in L-[14C]lactate transport into the control ( non-injected ) oocytes ( S4 Fig ) . Much higher concentrations of the compounds were required to inhibit L-[14C]lactate transport via PfFNTGly107Ser compared to those found to inhibit transport via native PfFNT ( Fig 4 ) . The IC50 values for the inhibition of PfFNT-mediated L-[14C]lactate transport by MMV007839 were 22 . 5 ± 1 . 4 nM ( mean ± SEM; n = 4 ) for native PfFNT and 1 . 3 ± 0 . 2 μM for PfFNTGly107Ser ( P < 0 . 001; one-way ANOVA with post hoc Tukey test ) . For MMV000972 , the IC50 values for native PfFNT and PfFNTGly107Ser were 50 ± 1 nM and 6 . 3 ± 0 . 9 μM , respectively ( P < 0 . 001 ) . These results provide direct evidence that PfFNT is inhibited by MMV007839 and MMV000972 and that the Gly107Ser mutation greatly reduces the sensitivity of PfFNT to inhibition by these compounds . The specificity of the inhibition by the MMV compounds of the transport of L-[14C]lactate via PfFNT was investigated by testing the effects of the compounds on [3H]chloroquine transport by oocytes expressing a chloroquine-transporting isoform of the P . falciparum chloroquine resistance transporter [26] ( PfCRT; from the chloroquine-resistant Dd2 strain ) and on [3H]hypoxanthine transport by oocytes expressing the P . falciparum equilibrative nucleoside/nucleobase transporter PfENT1 [27 , 28] . Neither compound had a statistically significant effect on the transport of [3H]chloroquine via PfCRT ( S5 Fig ) . Neither compound inhibited [3H]hypoxanthine uptake by PfENT1-expressing oocytes ( S5 Fig ) , although 1 μM MMV000972 caused a slight increase in [3H]hypoxanthine uptake by PfENT1-expressing oocytes ( P < 0 . 05; one-way ANOVA with post hoc Tukey test ) . The effect of MMV007839 ( the more potent of the two MMV compounds at inhibiting parasite growth ) on the metabolite profile of erythrocytes infected with mature 3D7 trophozoites was assessed using a previously-described untargeted LC-MS approach [29] . The compound was added at approximately 20× IC50 ( Table 1; 6 μM ) and samples were taken for metabolite profiling at 1 h , 3 h and 6 h after the addition . Exposure of 3D7 parasites to 6 μM MMV007839 for 1 h or 3 h did not significantly affect their viability ( P > 0 . 14; paired t-tests; S6 Fig ) , assessed by exposing parasitised erythrocytes to the inhibitor for the specified time , then removing the inhibitor and culturing the parasites for a further three days and measuring the final parasitaemia . By contrast , exposure for 6 h resulted in a reduction in parasitaemia ( as measured three days after the exposure ) to 49 ± 7% ( mean ± SEM; n = 3 ) of control levels ( P = 0 . 007; S6 Fig ) . The levels of a number of metabolites within infected erythrocytes were found to be affected by MMV007839 ( Fig 5; S1 Data ) . For those metabolites for which differences in abundance between MMV007839-treated and control samples were statistically significant , the relative levels ( i . e . the level of the metabolite in MMV007839-treated cells relative to that in untreated cells ) are shown in Fig 5B . Lactate was the metabolite that showed the greatest increase in abundance in MMV007839-treated infected erythrocytes relative to untreated infected erythrocytes ( a 6-fold increase at each time point; Fig 5A and 5B; S1 Data ) , consistent with the MMV compound inhibiting the efflux of lactate from the parasite . There was also a significant elevation of pyruvate ( the metabolite directly upstream of lactate ) , the glycolytic metabolite sn-glycerol-3-phosphate , and a number of peptides , in MMV007839-exposed cells ( Fig 5A and 5B ) . The levels of a number of metabolites were decreased in MMV007839-treated infected erythrocytes relative to untreated infected erythrocytes . These included NADH and the pyrimidine precursors orotate and orotidine-5-phosphate ( Fig 5A and 5B ) . The levels of a range of metabolites in the extracellular medium were also analysed ( Fig 5C ) . Extracellular lactate and α-ketoglutarate levels were decreased in MMV007839-treated samples relative to untreated samples , whereas the levels of pyruvate , phosphoenolpyruvate and 3-phosphoglycerate were increased . Phosphoenolpyruvate and 3-phosphoglycerate were the only metabolites for which differences in abundance between MMV007839-treated and untreated samples were statistically significant ( S1 Data ) . The MMV007839-induced accumulation of lactate ( and glycolytic intermediates ) within the parasite might be expected to induce cell swelling , through osmotic effects . To test this we investigated the effect of MMV007839 on the volume of both isolated ( 3D7 ) parasites and intact parasitised erythrocytes , using a Coulter Multisizer . Addition of MMV007839 ( 1 μM ) to a suspension of isolated parasites caused the parasites to swell , increasing their volume by 13 . 1 ± 2 . 8% ( relative to their starting volume; equating to a volume increase of 5 . 1 ± 0 . 8 fL; mean ± SEM; n = 3; P < 0 . 001 ) within 10 min and maintaining a similar volume for the next 30 min ( Fig 6A ) . MMV007839 ( 1 μM ) also caused erythrocytes infected with 3D7 trophozoites to swell , increasing their volume by 7 . 2 ± 1 . 3% ( relative to their starting volume; equating to a volume increase of 5 . 4 ± 1 . 0 fL; mean ± SEM; n = 3; P < 0 . 01 ) within 10 min and maintaining a similar volume for the next 50 min ( Fig 6B ) .
In this study we provide multiple lines of evidence that two structurally-related compounds from the MMV Malaria Box , MMV007839 and MMV000972 , kill asexual blood-stage P . falciparum parasites via inhibition of the lactate:H+ transporter PfFNT . Compounds with this mechanism of action have not been described or exploited as antimalarial drugs previously . Our findings highlight the potential of PfFNT as an antimalarial drug target . The most direct line of evidence that the two compounds target PfFNT comes from experiments in which PfFNT was studied in isolation from other P . falciparum transporters in the Xenopus oocyte . In this heterologous expression system , MMV007839 and MMV000972 inhibited L-[14C]lactate transport via the native form of PfFNT with IC50 values of 23 nM and 50 nM , respectively , while not inhibiting two other P . falciparum transporters ( PfCRT and PfENT1 ) . The IC50 value for PfFNT inhibition by MMV007839 is somewhat lower than the IC50 values we obtained for parasite growth inhibition by this compound ( 260 nM for 3D7 parasites and 140 nM for Dd2 parasites ) . The IC50 value for parasite growth inhibition by MMV000972 ( 1 . 8 μM for Dd2 parasites ) was much higher than that obtained for inhibition of native-PfFNT-mediated L-[14C]lactate transport . MMV007839 and MMV000972 have also been tested against 3D7 parasites by other groups , yielding IC50 values of 283–442 nM and 2 . 6 μM , respectively ( http://www . mmv . org/research-development/malaria-box-supporting-information ) . Given that MMV007839 and MMV000972 inhibit PfFNT with comparable potency when expressed in oocytes , the much higher IC50 value for parasite growth inhibition by MMV000972 compared to MMV007839 might result from a lower concentration of the former compound in the parasitised erythrocyte or the parasite itself . Parasites selected for resistance to growth-inhibition by MMV007839 had a Gly107Ser mutation in PfFNT . This mutation , which was observed in two independent cultures within six weeks of first adding the compound , decreased the susceptibility of parasites to MMV007839 and MMV000972 by > 30-fold while not affecting their sensitivity to chloroquine or artemisinin . The fact that the MMV007839-selected parasites were cross-resistant to MMV000972 was not surprising in light of the high degree of structural similarity between the compounds and our finding that they both inhibit PfFNT . The fact that we were able to generate parasites with high-level resistance to MMV007839 and MMV000972 within six weeks is a potential cause for concern with regard to the potential of such compounds as antimalarials . However , this would not necessarily deter further investigations into exploiting PfFNT as a drug target . Many compounds for which resistance has been generated in vitro are undergoing development [3] , such as the protein translation inhibitor DDD498 [30] and various compounds that have been proposed to target PfATP4 [31–33] . A relationship between the propensity of parasites to develop resistance in vitro and the likelihood of parasite resistance emerging in vivo is not firmly established . Factors such as the drug’s rapidity of action [34] and the growth rate and transmissibility of drug-resistant parasites also contribute to the overall threat of clinical resistance [35] . Furthermore , carefully selected dosing regimens and partner drugs may reduce the likelihood of selecting for resistance-conferring mutations in the target protein in vivo . Our direct studies of PfFNT inhibition were complemented by in situ studies with P . falciparum parasites . The parasite studies provided multiple lines of evidence ( based both on pH measurements and L-[14C]lactate flux measurements ) that both MMV compounds abolish L-lactate transport across the parasite plasma membrane . Experiments with L-[14C]lactate showed that MMV007839 and MMV000972 inhibited the transport of L-lactate across the parasite plasma membrane with IC50 values of 158 nM and 49 nM respectively in wild-type Dd2 parasites . These IC50 values are similar to those obtained for the inhibition of L-[14C]lactate uptake by PfFNT expressed in Xenopus oocytes . This suggests that PfFNT may be the only transporter that can mediate the transport of L-lactate across the parasite plasma membrane at an appreciable rate , at least in trophozoite-stage parasites . The parasite does encode two transporters with homology to human monocarboxylate transporters , however neither showed a capacity to transport lactate when expressed heterologously in yeast [17] . The potency with which MMV007839 and MMV000972 inhibited L-[14C]lactate transport by parasites was significantly reduced in the resistant parasites bearing the Gly107Ser mutation in PfFNT . Using the Xenopus oocyte system , this finding was attributed directly to a reduced inhibitor sensitivity of the mutant PfFNTGly107Ser protein compared to the native protein . The mutated residue is adjacent to T106 , which has been predicted to form part of the PfFNT transport channel [17] . The mutation may reduce the affinity with which MMV007839 and MMV000972 bind to the transporter , or otherwise prevent the compounds from inhibiting its function . The mutation was associated with an approximately 50–60% reduction in L-[14C]lactate transport by both parasites and PfFNT expressed in Xenopus oocytes . This reduction in lactate transport was not associated with a decrease in the growth rate of asexual blood-stage parasites , suggesting that the rate of lactate export from parasites must be inhibited by more than 50% to affect their viability under the conditions tested here . The metabolite profiles of MMV007839-treated and untreated parasitised erythrocytes provided further evidence in support of the mechanism of action proposed here . Lactate levels were specifically elevated in MMV007839-treated infected erythrocytes compared to untreated infected erythrocytes , consistent with decreased efflux from infected erythrocytes . The level of lactate secreted into the medium was also reduced in MMV007839-treated cultures relative to untreated cultures ( although this decrease did not reach statistical significance ) , while pyruvate , phosphoenolpyruvate and 3-phosphoglycerate secretion was increased in the MMV007839-treated infected erythrocytes . Collectively , these analyses indicate that MMV007839 induces a build-up of lactate , its immediate precursor ( pyruvate ) , and other glycolytic intermediates . This build-up is partly countered by increased secretion of partly oxidised intermediates and accompanied by significant swelling of both the parasite and the intact infected erythrocyte . The observed accumulation of lactate and pyruvate is a distinctive phenotype that has not been observed previously when infected erythrocytes were treated with a range of antimalarials including glycolysis inhibitors [29] . The metabolic perturbations were observed after 1 h , 3 h and 6 h exposures to MMV007839 . As parasite viability was not affected at the early time points , these changes cannot be attributed to cell death . NADH levels were significantly lower in MMV007839-treated compared to untreated infected erythrocytes . This might be explained by the consumption of NADH in the reaction that produces sn-glycerol 3-phosphate , which was significantly elevated within the MMV007839-treated cells . Significant changes in intracellular abundance were also observed for a range of other metabolites , including a number of ( likely haemoglobin-derived ) small peptides and pyrimidine precursors . The basis for these changes was not investigated , although they may be a secondary consequence of the MMV007839-induced acidification of the parasite cytosol or disruption to energy metabolism . Previous work has demonstrated that inhibiting the P . falciparum glucose transporter not only perturbs glycolysis but also impacts other metabolic pathways including pyrimidine biosynthesis and haemoglobin digestion [29] . Studies in which the fates of isotope-labelled metabolites are tracked would be required to resolve the mechanistic basis for the full range of metabolic perturbations resulting from inhibition of PfFNT . Thus while the changes in the levels of glycolytic intermediates observed in MMV007839-treated parasitised erythrocytes are fully consistent with inhibition of PfFNT , it remains possible that changes in the levels of other metabolites could have resulted from off-target effects of MMV007839 rather than as a consequence of downstream effects of PfFNT inhibition . A recently published study in which extensive toxicity testing was performed on the Malaria Box compounds revealed that MMV007839 and MMV000972 display toxicity in some assays [36] . The compounds themselves are therefore unlikely to be suitable as clinical candidates . Further screening for compounds that inhibit PfFNT will be required to develop suitable clinical candidates that might be used to exploit this target in the field . In summary , our screen of the Malaria Box for compounds that perturb the pH inside the parasite led to the discovery of two compounds that kill parasites via inhibition of a novel target , PfFNT . The mechanisms of action of the remaining pH-decreasing hits are yet to be determined , and further studies on these compounds may uncover additional targets . Another transport protein that mediates the translocation of H+ across the plasma membrane , PfATP4 , has emerged recently as a particularly vulnerable drug target [37] , and compounds believed to target this pump give rise to a cytosolic alkalinisation . Thus , pH assays can be used as a primary screen for the identification of additional compounds that inhibit either PfFNT or PfATP4 .
The Malaria Box [8] was provided by MMV . Upon receipt of the Malaria Box , each compound was diluted to 1 mM in DMSO and aliquoted into multiple plates . Information about the compounds is accessible via the ChEMBL-NTD database ( https://www . ebi . ac . uk/chemblntd ) . Additional quantities of MMV007839 and MMV000972 were obtained from MMV and Vitas-M Laboratory , respectively . The use of human blood in this study was approved by the Australian National University Human Research Ethics Committee . The blood was provided by the Australian Red Cross Blood Service without disclosing the identities of the donors . Ethical approval of the work performed with the Xenopus laevis frogs was obtained from the Australian National University Animal Experimentation Ethics Committee ( Animal Ethics Protocol Number A2013/13 ) in accordance with the Australian Code of Practice for the Care and Use of Animals for Scientific Purposes . The 3D7 strain of P . falciparum ( isolated in the Netherlands but likely to be of African origin ) was used throughout this study , except for the MMV007839 resistance selection studies , which were initiated with a clone of Dd2 ( Thai origin ) that had been generated previously by limiting dilution [38] , and the subsequent experiments performed to characterise the resistant parasites . The 3D7 strain and the Dd2 clone were checked for mycoplasma contamination prior to their use in this study and were found to be mycoplasma-free . Parasites were cultured in human erythrocytes [39] , with continuous shaking [40] , and were synchronised by sorbitol treatment [41] . The culture medium was RPMI 1640 containing 25 mM HEPES ( Gibco ) supplemented with 11 mM additional glucose , 0 . 2 mM hypoxanthine , 20 μg/mL gentamicin sulphate and 3 g/L Albumax II . Prior to cytosolic pH , digestive vacuole pH , parasite volume and L-[14C]lactate uptake measurements , mature trophozoite-stage parasites ( approximately 34–40 h post-invasion ) were functionally isolated from their host erythrocytes by brief exposure ( of cultures at approximately 4% haematocrit ) to saponin ( 0 . 05% w/v , of which ≥ 10% was the active agent sapogenin ) [42] . Unless stated otherwise the parasites were then washed several times in bicarbonate-free RPMI 1640 supplemented with 11 mM additional glucose , 0 . 2 mM hypoxanthine and 25 mM HEPES ( pH 7 . 10 ) , and maintained in this medium at a density of 1 × 107–3 × 107 parasites mL-1 at 37°C until immediately before their use in experiments . Cytosolic pH was measured as described previously [25] at 37°C ( unless stated otherwise ) using saponin-isolated trophozoites loaded with the pH-sensitive fluorescent dye BCECF ( Molecular Probes ) . In most experiments parasites were suspended in ‘Experimental Saline Solution’ ( 125 mM NaCl , 5 mM KCl , 1 mM MgCl2 , 20 mM glucose , 25 mM HEPES; pH 7 . 10 unless stated otherwise ) . For experiments in which parasites were deprived of glucose , glucose-free Experimental Saline Solution was used ( 135 mM NaCl , 5 mM KCl , 1 mM MgCl2 , 25 mM HEPES; pH 7 . 10 ) . Parasites suspended at a density of 1 × 107–3 × 107 parasites mL-1 were excited successively at 440 and 495 nm , with emission recorded at 520 nm , using either a PerkinElmer LS 50B Fluorescence Spectrometer or a Cary Eclipse Fluorescence Spectrophotometer . The relationship between the ratio of the two measurements ( 495 nm/440 nm ) and the cytosolic pH was calibrated as described previously [25] with one exception: for experiments performed at 4°C , the previously published calibration method yielded abnormally high values for resting pH , and the calibration was therefore performed using Experimental Saline Solution ( adjusted to various pH values ) in place of the high-K+ calibration solutions . The pH inside the digestive vacuole ( pHDV ) was monitored at 37°C using isolated parasites containing fluorescein-dextran ( 10 , 000 MW; Molecular Probes ) in their digestive vacuoles . The parasites were prepared as described previously [43] and suspended in Experimental Saline Solution ( pH 7 . 10 ) at a density of 1 × 107–3 × 107 parasites mL-1 . Fluorescence ratio measurements were obtained with a PerkinElmer Life Sciences LS 50B fluorometer with a dual excitation Fast Filter accessory using excitation wavelengths of 440 and 495 nm and an emission wavelength of 520 nm [44] . The uptake of L-[14C]lactate ( PerkinElmer ) by isolated 3D7 trophozoites was determined essentially as described previously [16] . Experiments were performed at 4°C on parasites suspended in pH 6 . 10 Experimental Saline Solution . Parasites were preincubated with either DMSO or inhibitors ( at the concentrations stated in the relevant Figure legends ) for 1 min before a 200 μL sample of the suspension was added ( at ‘time zero’ ) to an equal volume of pH 6 . 10 saline ( with or without inhibitors as specified in the Figure legends ) layered over a 200 μL dibutyl phthalate/dioctyl phthalate ( 5:4 v/v ) oil mix ( in a microcentrifuge tube ) and containing L-[14C]lactate ( giving a final concentration after the cell suspension was added of 1 . 3 μM ) . At 20 s after combining the parasites and L-[14C]lactate , uptake was terminated by centrifuging the sample ( 15 , 800 × g , 2 min in a rapid-acceleration Beckman Microfuge E ) , thereby sedimenting the parasites below the oil layer . A 10 μL aliquot of the supernatant solution was taken from above the oil layer in each tube to enable the determination of extracellular L-[14C]lactate concentrations . The remaining supernatant solution was discarded and residual radioactivity on the sides of each tube was removed by rinsing the tube four times with water before aspirating most of the oil . The cell pellets were lysed using Triton X-100 ( 0 . 1% v/v ) and deproteinised with trichloroacetic acid ( 2 . 5% w/v final ) before the radioactivity in each sample was measured using a scintillation counter [42] . In each experiment the amount of radioactivity in the cell pellets that was attributable to ‘trapped’ extracellular L-[14C]lactate was estimated under conditions in which the amount of L-[14C]lactate taken up by the parasites was minimised by using the previously-described [16] lactate transport inhibitor NPPB . Parasites were preincubated with NPPB ( 100 μM ) for 1 min then combined with a L-[14C]lactate solution to which unlabelled L-lactate and NPPB had been added ( yielding a final lactate concentration of 10 mM and a final NPPB concentration of 100 μM ) , then centrifuged immediately . ‘Distribution ratios’ ( the intracellular L-[14C]lactate concentration divided by the extracellular L-[14C]lactate concentration ) were calculated as described previously [16] . The extracellular L-[14C]lactate concentration for each sample was determined using the aliquots of supernatant solution taken from above the oil layer . The intracellular L-[14C]lactate concentration for each sample was determined from the radioactivity incorporated into the cell pellet ( after subtraction of the radioactivity attributable to the trapped extracellular L-[14C]lactate ) , the parasite number , and the previously determined estimate of the water volume of a saponin-isolated parasite [42] . Parasite proliferation was measured in 96-well plates using a fluorescent DNA-intercalating dye [45] . The assays were initiated with erythrocytes infected with predominantly ring-stage parasites , and the starting parasitaemia was 0 . 5–1% and the haematocrit was 1% . The experimental protocol and procedure for data analysis were the same as those outlined previously [46] with two modifications: the duration of the assay was 68–72 h , and a supramaximal concentration of an antiplasmodial compound ( either 5 μM chloroquine , 0 . 5 μM artemisinin , 12 . 8 nM KAE609 or 150 μM MMV007839 ) was used for the ‘zero growth’ control . The concentration of DMSO in the assays did not exceed 0 . 2% ( v/v ) . The volume of saponin-isolated 3D7 trophozoites and of erythrocytes infected with 3D7 trophozoites was measured using a Beckman Coulter Multisizer 4 fitted with a 100 μm ‘aperture tube’ . In the case of isolated parasites , the cells were washed and resuspended ( at 37°C ) in pH 7 . 10 Experimental Saline Solution . The electrolyte solution within the aperture tube was the same as the Experimental Saline Solution except that it lacked glucose . For experiments on infected erythrocytes , the infected cells were separated from uninfected erythrocytes using a Miltenyi Biotec VarioMACS Magnet [47 , 48] and the cells were maintained at 37°C in bicarbonate-free RPMI supplemented with 25 mM HEPES , additional glucose ( 11 mM ) and 0 . 2 mM hypoxanthine , and pH-adjusted to 7 . 40 . For these measurements the electrolyte solution within the aperture tube differed only from this medium in that it was not supplemented with hypoxanthine or additional glucose . For each measurement of cell volume , approximately 20 , 000 pulses ( each corresponding to the passage of a single cell through the aperture ) were recorded . The median volume of the cells within each sample was determined by fitting a log Gaussian distribution curve to the population data . A clone of the chloroquine-resistant Dd2 strain of P . falciparum was used to generate MMV007839-resistant parasites . This clone was generated previously by limiting dilution and was used successfully to generate resistance to two ( PfATP4-associated ) Malaria Box compounds [18] . Two independent cultures ( each containing ~ 5 × 108 parasites ) were exposed to increasing concentrations of MMV007839 for ~ 6 weeks , starting with a concentration of 100 nM on Day 0 , and increasing to 125 nM ( approximately the IC50 value for the parental Dd2 line ) on Day 3 . For MMV007839-selected culture A , the drug concentration was increased by 5 nM increments on Day 18 , Day 21 , and every two days thereafter . For MMV007839-selected culture B , the drug concentration was increased by 15 nM increments every two days starting on Day 18 . The cultures were provided with fresh medium and blood and diluted to reduce the parasitaemia as needed . Genomic DNA was extracted from saponin-isolated parasites ( from the MMV007839-selected bulk cultures ) using a QIAGEN Plant DNeasy kit . To sequence exons 1–3 , the entire PfFNT gene was PCR-amplified using KOD Hot Start DNA polymerase and primers 1 and 2 ( S2 Table ) . The PCR product was extracted using a GeneJET Gel Extraction Kit ( Thermo Fisher Scientific ) . Sequencing was performed at the ACRF Biomolecular Resource Facility ( The John Curtin School of Medical Research , Australian National University ) using primers 1 and 3–7 ( S2 Table ) . To sequence the fourth ( and final ) exon , the 3’ end of the gene and part of the 3’ UTR was PCR amplified using the primers 3 and 8 ( S2 Table ) , with primer 8 also used for sequencing . For fitness assays , two clones were obtained from MMV007839-selected culture B by limiting dilution , performed essentially as described previously [38] . The presence of the G319A mutation in the PfFNT coding sequence ( giving rise to the Gly107Ser mutation in PfFNT ) was confirmed in both clones . Erythrocytes infected with 3D7 trophozoites were separated from uninfected erythrocytes using a magnet supplied by Colebrook Bioscience [47 , 48] . The resulting cells , which consisted of > 95% infected erythrocytes , were then allowed to recover for 0 . 5–1 h at 37°C in ‘complete medium’ ( RPMI 1640 supplemented with 0 . 5% Albumax II , 25 mM HEPES , 100 μM hypoxanthine and 10 μg/mL gentamycin ) . At ‘time zero’ , MMV007839 ( 6 μM ) was added to three cell suspensions ( each containing 1 × 108 cells at 0 . 4% haematocrit ) , while three identical cell suspensions were left untreated . The cell suspensions were incubated at 37°C under controlled atmospheric conditions ( 5% CO2 and 1% O2 in N2 ) . At each of three time points ( 1 h , 3 h and 6 h ) , one MMV007839-treated cell suspension and one untreated cell suspension were used to extract metabolites . To extract metabolites , cell suspensions were first pelleted by centrifugation ( 14 , 000 × g , 30 s ) . Aliquots ( 5 μL ) of the supernatant media were collected and the remaining supernatant media discarded . The cells were then washed with 1 mL of ice-cold PBS . To extract metabolites , 200 μL volumes of 80% acetonitrile ( in H2O; containing 5 μM [13C]aspartate as an internal standard ) were added to the cell pellets and samples of extracellular media . The samples were vortexed briefly then centrifuged ( 14 , 000 × g , 10 min , 4°C ) , and the resulting supernatant solutions were transferred to vials for LC-MS analysis . The metabolites were separated on a SeQuant ZIC-pHILIC column ( 5 μm , 150 × 4 . 6 mm , Millipore ) with a 1200 series HPLC system ( Agilent ) , using a flow rate of 0 . 3 mL/min with 20 mM ammonium carbonate in water ( A ) and 100% acetonitrile ( B ) as the mobile phase . A binary gradient was set up as follows: 0 . 5 min: 20% A and 80% B , 30 min: 80% A and 20% B , 31 min: 95% A and 5% B , 35 . 5–45 min: 20% A and 80% B . Detection of metabolites was performed on an Agilent Q-TOF mass spectrometer 6550 operating in negative ESI mode . The scan range was 85–1200 m/z between 5 and 35 min at 0 . 9 spectra/second . LC-MS . d files were converted to . mzXML files using MS convert and analysed using MAVEN [49] . Following alignment , metabolites were assigned using exact mass ( < 5 ppm ) and retention time ( compared to a standards library of 150 compounds run on the same day ) . The area top for each positively assigned metabolite was integrated and used to calculate the ratio of the metabolite concentration between MMV007839-treated and untreated controls . Log2 ratios ( + MMV007839/- MMV007839 ) across the time series were then plotted using the heatmap script in R . The Xenopus laevis oocyte expression system was used to assess the effects of MMV007839 and MMV000972 on PfFNT ( from the 3D7 strain ) , PfENT1 ( from the FAF6 strain ) and PfCRT ( from the Dd2 strain ) . The oocyte expression vectors containing PfFNT , PfENT1 and PfCRT were made previously [16 , 26 , 50] . The Gly107Ser mutant form of PfFNT was generated through site-directed mutagenesis of the oocyte expression vector containing wild-type ( 3D7 ) PfFNT using primers 9 and 10 ( S2 Table ) . Oocytes were harvested from adult female Xenopus laevis frogs and prepared as described previously [51] . cRNA was made using the mMessage mMachine T7 transcription kit and the MEGAclear kit ( Ambion ) and microinjected into oocytes as outlined elsewhere [51] . The amount of cRNA injected ( per oocyte ) was 30 ng for PfFNT , 10 ng for PfENT1 , and 20 ng for PfCRT . The uptake of radiolabelled substrates was measured 2–5 days post-injection at 27 . 5°C in ND96 buffer ( containing 96 mM NaCl , 2 mM KCl , 1 mM MgCl2 , 1 . 8 mM CaCl2 , 10 mM MES and 10 mM Tris-base; pH-adjusted to the value specified in the relevant Figure legends ) . L-[14C]lactic acid ( Na+ salt; 150 . 6 mCi/mmol ) and [3H]hypoxanthine monochloride ( 14 Ci/mmol ) were purchased from Perkin Elmer , and [3H]chloroquine ( 20 Ci/mmol ) was purchased from American Radiolabeled Chemicals . Each experiment used ten oocytes for each condition tested . The influx of the radiolabelled substrate was halted by removing the reaction buffer and washing the oocytes twice in ice-cold ND96 buffer ( 3 . 5 mL ) . The oocytes were then transferred to separate wells of a 96-well plate , lysed with 10% SDS , and the radioactivity measured as described previously [52] . For measurements of cell volume , one-way ANOVAs were carried out with the pre-normalised data using ‘experiment’ as a ‘blocking factor’ , to prevent differences in the starting volume between independent experiments ( which may have resulted from differences in the average age of the parasites on the different days ) from eroding the precision of the test . Post hoc comparisons were then performed using the least significant difference test . For the LC-MS data , two-way ANOVAs were performed for both the extracellular and intracellular metabolite datasets . Between-subject testing was performed and the results corrected for multiple hypothesis testing using the False Discovery Rate determined by MetaboAnalyst [53] . For all other comparisons , P values were obtained using either one-way ANOVAs followed by post hoc Tukey tests or t-tests ( paired or unpaired as appropriate ) , as stated in the relevant sections . All tests were two-sided . | The emergence and spread of Plasmodium falciparum strains resistant to leading antimalarial drugs has intensified the need to discover and develop drugs that kill the parasite via new mechanisms . Here we screened compounds that are known to inhibit P . falciparum growth for their effects on the pH inside the parasite . We identified fifteen compounds that decrease the pH inside the parasite , and determined the mechanism by which two of these , MMV007839 and MMV000972 , disrupt pH and kill the parasite . The two compounds were found to inhibit the P . falciparum formate nitrite transporter ( PfFNT ) , a transport protein that is located on the parasite surface and that serves to remove the waste product lactic acid from the parasite . The compounds inhibited both the H+-coupled transport of lactate across the parasite plasma membrane and the transport of lactate by PfFNT expressed in Xenopus oocytes . In addition to disrupting pH , PfFNT inhibition led to a build-up of lactate in the parasite-infected red blood cell and the swelling of both the parasite and the infected red blood cell . Exposing parasites to MMV007839 over a prolonged time period gave rise to resistant parasites with a mutant form of PfFNT that was less sensitive to the compound . This study validates PfFNT as a novel antimalarial drug target . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"blood",
"cells",
"medicine",
"and",
"health",
"sciences",
"parasite",
"groups",
"plasmodium",
"tropical",
"diseases",
"vertebrates",
"parasitic",
"diseases",
"parasitic",
"protozoans",
"animals",
"parasitology",
"xenopus",
"animal",
"models",
"apicomplexa",
"model",
"organisms",
"amphibians",
"protozoans",
"red",
"blood",
"cells",
"metabolites",
"experimental",
"organism",
"systems",
"bioassays",
"and",
"physiological",
"analysis",
"research",
"and",
"analysis",
"methods",
"transport",
"inhibition",
"assay",
"malarial",
"parasites",
"animal",
"cells",
"xenopus",
"oocytes",
"biochemistry",
"cell",
"biology",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"malaria",
"metabolism",
"frogs",
"organisms"
] | 2017 | The Malaria Parasite's Lactate Transporter PfFNT Is the Target of Antiplasmodial Compounds Identified in Whole Cell Phenotypic Screens |
Meiotic recombination is a fundamental cellular process , with important consequences for evolution and genome integrity . However , we know little about how recombination rates vary across the genomes of most species and the molecular and evolutionary determinants of this variation . The honeybee , Apis mellifera , has extremely high rates of meiotic recombination , although the evolutionary causes and consequences of this are unclear . Here we use patterns of linkage disequilibrium in whole genome resequencing data from 30 diploid honeybees to construct a fine-scale map of rates of crossing over in the genome . We find that , in contrast to vertebrate genomes , the recombination landscape is not strongly punctate . Crossover rates strongly correlate with levels of genetic variation , but not divergence , which indicates a pervasive impact of selection on the genome . Germ-line methylated genes have reduced crossover rate , which could indicate a role of methylation in suppressing recombination . Controlling for the effects of methylation , we do not infer a strong association between gene expression patterns and recombination . The site frequency spectrum is strongly skewed from neutral expectations in honeybees: rare variants are dominated by AT-biased mutations , whereas GC-biased mutations are found at higher frequencies , indicative of a major influence of GC-biased gene conversion ( gBGC ) , which we infer to generate an allele fixation bias 5 – 50 times the genomic average estimated in humans . We uncover further evidence that this repair bias specifically affects transitions and favours fixation of CpG sites . Recombination , via gBGC , therefore appears to have profound consequences on genome evolution in honeybees and interferes with the process of natural selection . These findings have important implications for our understanding of the forces driving molecular evolution .
In most sexual eukaryotes , average recombination rates do not greatly exceed one crossover per chromosome arm , which is commonly a minimum requirement for correct meiosis [1] . However , the honeybee , Apis mellifera , has extremely high recombination rates , averaging 19–37 cM/Mb [2–4] , which corresponds to more than 5 crossovers per chromosome pair per meiosis . Such high rates are observed in other social insects but not their solitary cousins [5 , 6] . This suggests that high recombination rates are an adaptation favoured by eusociality although the specific causes are unknown . Insight into this question can be gained by analysing the fine-scale landscape of recombination rate variation in order to understand the molecular mechanisms that govern it . The molecular mechanisms that determine the genomic distribution of recombination events in honeybees are unclear . In a wide range of species , recombination events are strongly clustered into short hotspots a few kb in length [7–10] . In human and mouse , these hotspots are found to be enriched for a DNA motif recognised by the protein PRDM9 [11–14] . This protein binds to the DNA motif and catalyses a histone modification that acts as a mark for the formation of a DNA double stranded break in the same location [15] . In species without an active PRDM9 , hotspots are often present , but other features may define them . For example , in dog , where PRDM9 is inactive , recombination events are clustered in un-methylated CpG islands [16–18] . In yeast and Arabidopsis recombination hotspots are observed in nucleosome-depleted open chromatin and gene promoters [8 , 9] . The few invertebrate genomes analysed so far tend to lack extreme recombination hotspots [19 , 20] . In particular , recombination rates in the fruit fly Drosophila melanogaster appear to be less variable across the genome than other species where fine-scale genetic maps are available [20–23] . Genetic maps of the honeybee do not indicate the presence of hotspots with extremely elevated rates [2 , 3 , 24] or the presence of enriched sequence motifs [4] . This is consistent with the absence of a PRDM9-like mechanism controlling recombination rates in insects and suggests that other factors are more important . One such factor could be DNA methylation . Unlike fruit flies , the honeybee has an intact methylation system [25 , 26] . It is therefore possible that rates of recombination in the honeybee genome are influenced by DNA methylation patterns , as observed in some other taxa [16–18 , 27] . In a diverse range of species , local rates of crossing-over correlate with genetic diversity but not with genetic divergence [28 , 29] . These correlations are inferred to be due to an indirect effect of recombination due to the interaction between selection and linkage and their strength can be used to make inferences about the pervasiveness of natural selection . Positive selection on favourable mutations or negative selection against deleterious changes reduce levels of linked variation by the processes of genetic hitchhiking and background selection and these effects are predicted to be larger in regions of low recombination [30 , 31] , resulting in lower genetic diversity in these regions . Strong correlations exist in many species of fruit fly that have been used to predict that large proportions of the genome are affected by selection [32–34] , whereas in humans such correlations are weaker [35 , 36] , suggesting a less pervasive impact of selection on genetic variation . Social insects such as honeybees have lower effective population sizes than solitary ones [37 , 38] and it is unclear if selection has a similarly pervasive impact on genome variation . A number of hypotheses have been proposed to explain the extremely high recombination rates in honeybees and other social insects . One class of hypotheses suggests that they represent an adaptation important for the evolution of behavioural phenotypes in the worker caste . This could be because the evolution of eusociality entailed rapid evolution and specialisation of workers [39] . Alternatively , high intra-colony variability in worker phenotypes could be beneficial because it results in a more efficient workforce [40 , 41] . These factors could lead to increased recombination rates in the vicinity of genes specifically involved in worker phenotypes . Some studies have reported evidence for higher recombination rates in genes with worker-biased expression [4 , 39] . However , the cause of these associations is unclear and several questions remain . In particular , it is not known whether worker-biased genes are preferentially located in regions of high recombination , or whether there is a direct influence of gene expression or a related process on recombination rate within genes . Recombination can have profound affects of genome evolution via GC-biased gene conversion ( gBGC; reviewed in [42] ) . This process is believed to occur due to the biased repair of nucleotide mismatches that occur in heteroduplex DNA generated from pairing of two alleles during meiotic recombination . This involves a small bias towards repairing a mismatch involving a G/C ( or S , for strong ) nucleotide paired with an A/T ( or W , for weak ) nucleotide in favour of retaining the S allele , which results in an increased probability of transmitting the S allele into the gametes . There is a large amount of indirect evidence that this process occurs , indicating that genomic regions of high recombination accumulate GC-biased nucleotide substitutions over evolutionary time [43 , 44] , which results in a correlation between recombination and GC content [45] . A transmission bias towards S alleles has also been directly observed in yeast [46] by analysis of the products of meiosis and humans [47] by analysing transmission through pedigrees . The population dynamics of gBGC are equivalent to selection acting to increase the fixation probability of weak-to-strong ( WS ) mutations [48] . As such , gBGC can have effects on the site frequency spectrum [49–51] and rate of nucleotide substitution [52–54] similar to selection . It can also interfere with the process of natural selection . For example , gBGC could cause increased substitution rates in functional regions that can be mistaken for positive selection [52–54] . It can also lead to fixation of deleterious changes , including those underlying genetic disease in humans [55 , 56] . The transmission bias caused by gBGC also results in a skewed allele frequency spectrum , where WS mutations segregate at higher frequencies . Glémin et al . [57] modelled this property to estimate the strength of gBGC in the human population . The average strength of gBGC , B , was estimated as 0 . 38 , ( where B = 4NEb , NE being the effective population size and b the gBGC coefficient ) but 1% to 2% of the genome was estimated to be subject to strong gBGC with B>5 . Phylogenetic estimates indicate variation in B over two orders of magnitude among placental mammals [58] . The extreme recombination rates of honeybees could also indicate that gBGC is also very powerful , suggesting it could significantly impact molecular evolution in honeybees . In support of this previous studies found elevated frequencies of WS mutations , particularly in regions of high GC content [2 , 39] . A striking and unique feature of the honeybee genome is the over-representation of CpG dinucleotides [26] . The statistic CpGO/E measures the frequency of CpG dinucleotides in a nucleotide sequence compared to its expected value based on individual frequencies of Cs and Gs . In species where most CpG sites in the genome are methylated , as is the case in plants and vertebrates , CpG sites occur at a much lower frequency than expected due to the effects of methylated CpG hypermutability ( average CpGO/E in humans is 0 . 2 ) and this value rarely exceeds one in eukaryote genomes . The honeybee genome is unique in that it has a much higher frequency of CpG sites than expected ( CpGO/E is around 1 . 67 ) . The reason for this is unclear , but possible explanations are a mutational bias in favour of CpG sites or a fixation bias due to gBGC that favours the fixation of mutations that generate CpG dinucleotides . There are a number of unresolved questions regarding the evolution , molecular control and consequences of recombination in the honeybee genome . Firstly , is there evidence for recombination hotspots ? How does gene expression and DNA methylation affect local rates of recombination ? The answer to these questions could give us insight into how recombination is controlled in invertebrates . Secondly , does recombination modulate strength of natural selection across the genome ? This can be addressed by investigating the correlation between recombination rate and the levels of genetic diversity and divergence . Thirdly , is there evidence for a local increase in recombination rate in the vicinity of genes with worker-biased expression ? It has been suggested that this could be selectively advantageous due to the importance of worker phenotypes in the evolution of eusociality . Finally , what effects do the extremely high levels of recombination in honeybee have on the strength of gBGC ? How does gBGC impact genome variation and the frequency of CpG sites in the genome ? We can address these questions by analysing the shape of the site frequency spectrum for different SNP categories and estimating the value of B . Here we construct a fine-scale map of recombination rate variation honeybee using population-scale resequencing dataset [37] with the aim of addressing these questions . Our estimates show good correspondence with a previous genetic map [3] . Compared to the human genome , recombination events do not appear to be strongly partitioned into hotspots in the honeybee genome . Our data is consistent with an effect of germline methylation generating variation in crossover rate by suppressing recombination . We find evidence for a strong association between recombination and levels of genetic variation . In contrast to previous studies , we do not find that worker-biased expression is a strong predictor of high recombination rate compared to other factors . We also uncover a major effect of recombination on genome variation via the process of gBGC , which is stronger than observed in any other species and has a major impact on genome variation and evolution .
We constructed a high-resolution map of rates of crossing-over from patterns of linkage disequilibrium among 6 . 2 million SNPs observed across 60 copies of the sixteen nuclear honeybee chromosomes . We chose to use samples from Africa , sequenced as part of a larger study , because they are the most genetically diverse and because there is no evidence for population structure between them [37] . We used the LDhat method , which estimates the population-scaled recombination rate ρ across the genome . This is related to the recombination frequency r by the equation ρ = 3NEr in the case of haplodiploid species , where NE is effective population size . The LD map contained 306 , 764 discrete rate intervals . About 50% of the genome is covered by intervals of 5 kb or longer in this map . The mean recombination rate is 390 ρ/kb and the average rate change is ~9% between adjacent intervals . Scaling by a NE of 500 , 000 , estimated previously using the same set of African samples [37] , this corresponds to an average crossover frequency , r , of 26 . 0 cM/Mb , which is in agreement with previous estimates of 19–37 cM/Mb [2–4] . Recombination does not appear to be strongly restricted to a limited portion of the genome ( Fig 1A ) , suggesting that there are not strong hotspots in honeybee genome but a relatively continual recombination landscape . For example 50% of the recombination events in the genome occur in 32% of the genome . In humans , a similar map from population scale sequencing suggests that 50% occurs in less than 10% of sequence [59] . There is however considerable large-scale variation in recombination rates along the chromosomes ( Fig 1B shows variation along Group 1 ) . The mean population-scaled recombination rate computed from 100 kb windows is 385 ρ/kb , with a standard deviation of 167 ρ/kb ( see S1 Fig for LD maps of all chromosomes ) . We find that the LD map is broadly congruent with a previously constructed genetic map [3] ( e . g . R2 = 0 . 341 for the large metacentric chromosome Group1; Fig 1B and 1C ) , but the strength of correlations varies among regions and chromosomes ( R2 = 0 . 213 across the genome; Figs 1D and S1 ) . There is a highly significant correlation between levels of neutral genetic diversity , measured by Watterson’s theta , θw , estimated using noncoding sites , and rates of crossing over in the honeybee genome ( R2 = 0 . 615 , Fig 2A ) . We also examined the relationship between crossing over rates and divergence between A . mellifera and A . cerana and found a significant but very weak correlation ( R2 = 0 . 018 , Fig 2B ) . The strong correlation between recombination and genetic variation remains after correcting for divergence ( R2 = 0 . 617 , Fig 2C ) . These correlations are also found separately in intronic , intergenic and coding regions ( S2 Fig ) . A highly significant but weaker correlation between diversity/divergence is found using average pairwise heterozygosity ( π ) to measure genetic diversity instead of θw ( R2 = 0 . 480 ) . A correlation between genetic variation corrected for divergence and recombination rate is consistent with the pervasive influence of linked selection on patterns of variation , due to background selection , recurrent selective sweeps , or both [28] . It is also possible that fixation biases due to GC-biased gene conversion ( gBGC ) could contribute to the correlation between genetic diversity and crossover rate , as the strength of gBGC is expected to covary with recombination [42] . To examine this possibility , we recomputed diversity ( using θw ) while removing large classes of non-coding A . mellifera SNPs and substitutions between A . mellifera and A . cerana that may putatively be affected by gBGC . We first removed all variants that change GC content and found diversity to still be positively correlated with crossover rate ( R2 = 0 . 563 ) . After observing gBGC among transitions in particular ( see below ) , we next removed all transitions and also observed a positive correlation between diversity and recombination ( R2 = 0 . 586 ) . These patterns favour linked selection as a major force in shaping variation in the genome . These correlations are only slightly weaker than the correlations observed when the dataset is randomly subsampled to the same size ( R2 = 0 . 573 and R2 = 0 . 594 respectively ) , which suggests that gBGC has at most a small effect on determining the magnitude of genetic variation in a genomic region . Average Tajima’s D is negative ( -1 . 178 ) reflecting of skew towards rare variants , as already observed in this African honeybee population [37] , which is indicative of population expansion . Tajima’s D ( measured in 100 kb windows ) shows a weak negative correlation with both GC content ( R2 = 0 . 114 ) and recombination rate ( R2 = 0 . 015 ) , which indicates a slightly higher skew towards rare variants in regions of high recombination . Pervasive linked selection is expected to generate a skew towards rare variants in regions of low recombination [32] , which we do not observe . This could indicate the action of additional factors . In order to assess whether the association between genetic diversity and inferred recombination rates could be an artefact of having more power to detect recombination in regions of high genetic variation , or due to other biases , we estimated LD-based maps of recombination using datasets where SNPs were removed or using different parameter as follows: i ) we produced a dataset where genetic variation ( θw ) in each 100 kb window was capped at 0 . 002 , effectively subsampling data in 98% of 100 kb windows; ii ) we produced a dataset where rare variants ( minor allele frequency<0 . 1 ) were removed; iii ) we evaluated the effect of increase the block penalty to 10 , which affects the probability of changes in recombination rate between genomic regions . In each case , the resulting LD maps were strongly correlated with the original map ( S3 Fig ) . The strong correlation between levels of genetic variation ( in the original dataset ) and inferred rates of crossing over remained in the LD maps produced using different parameters ( S3 Fig ) and are similar to that observed in the original dataset . We therefore conclude that variation in genetic variation across the honeybee genome does not generate biases in inference of recombination and that the correlation between recombination and genetic variation is real . Levels of genetic variation are reduced close to genes , indicative of an effect of linked selection [37] . In order to determine the effects of linked selection acting on coding sequence on our observed correlation between genetic variation and recombination , we analysed this correlation restricting the analysis to sites at different distances from genes . We find that the correlation between diversity and recombination in intergenic regions is weaker when restricted to sites far from coding sequences . At sites <20 kb from coding sequences R2 is 0 . 364 , whereas it is 0 . 281 at 50–60 kb and only 0 . 208 at 100–110 kb away ( randomly sampling the same amount of data in each case ) . This supports the interpretation that this correlation is due to the effect of linked selection , as sites under selection are expected to be rarer far from genes . This is also supported by a finding of an excess of SNPs with high FST in within coding sequences [37] . The honeybee genome has low GC content ( on average 34% ) but GC content is variable across the genome [26] . To further understand the basis for this variation , and how it relates to recombination rate variation , we first partitioned the genome according to annotations and calculated the average GC content among genes and gene elements . We find that coding , intergenic and intronic regions each have characteristic GC content ( S4 Fig ) . Coding regions are biased toward high GC content ( 39% ) , whereas intronic regions are particularly low ( 23% ) and intergenic regions are intermediate ( 31% ) . Interestingly , 5’ UTRs have higher on average GC content than 3’ UTRs ( 31% cf . 24% ) . A unique feature of the honeybee genome is an overall excess of CpG dinucleotides ( measured by CpGO/E ) , compared to expectations based on frequencies of single bases . CpGO/E is also highly variable between different functional regions of the genome . It is high in noncoding regions ( ~1 . 7 in both introns and intergenic regions ) . However , notably , the coding part of the genome has an average CpGO/E close to the expected ( 1 . 04 ) . This is consistent with the observation that methylation in honeybees occurs predominantly in gene bodies [60] . As reported previously [61] , CpGO/E is bimodally distributed among genes ( S4 Fig ) . We assigned genes into high or low CpGO/E categories compared to the mean of 1 . 19 . Genes with low average CpGO/E ( LCpG; <1 . 19 CpGO/E ) have high levels of germline methylation at CpG sites and tend to be associated with cellular housekeeping functions , whereas genes in the higher average CpGO/E class ( HCpG; >1 . 19 CpGO/E ) have low levels of germline methylation and tend to be caste and tissue specific [37 , 60–62] . We find a strong correlation between crossover rates and GC content in the honeybee genome ( R2 = 0 . 436 ) . Strong correlations are also observed between GC content and crossover rates within coding ( R2 = 0 . 506 ) , intronic ( R2 = 0 . 463 ) and intergenic regions ( R2 = 0 . 446; Fig 3A ) . These correlations are also observed in 5’ and 3’ UTRs ( S5 Fig ) . Such , correlations between GC content and crossover rates are observed in a wide variety of taxa , and could suggest that recombination drives GC content via the process of gBGC [42] . We find that CpGO/E is correlated with recombination in coding sequence ( R2 = 0 . 369 ) but only very weakly correlated in intronic ( R2 = 0 . 066 ) and intergenic regions ( R2 = 0 . 011; Fig 3B ) . Methylation is mainly restricted to coding sequence in honeybees and variation in CpGO/E in coding sequence is likely to reflect differences in germline methylation [60–62] . Conversely , variation in CpGO/E in other parts of the genome is not influenced by DNA methylation and also does not correlate with recombination rate . These results may therefore suggest a role of germline DNA methylation in attenuating recombination rates in the honeybee genome . Interestingly , we also detect this correlation in 3’ UTRs ( R2 = 0 . 289 ) , but not in 5’ UTRs , ( R2 = 0 . 025; S5 Fig ) which could indicate an effect of differential levels of methylation . In support of this , the CpGO/E distribution of 3’ UTRs is shifted towards lower CpGO/E compared to noncoding regions , indicative of higher levels of DNA methylation ( S4 Fig; [61] ) . Average rates of crossing over are reduced in coding sequence and UTRs compared to noncoding regions ( Fig 4A ) . This suggests the presence of specific factors that reduce recombination specifically within genes . We next examined how variation in patterns of gene expression and inferred levels of germline methylation are associated with crossover rate in genes ( Fig 4B ) . Previous studies have suggested that genes with worker-biased expression tend to have high recombination rates [4 , 39] . To test this , we first compared rates of crossing over within genes with biased expression in queens compared to workers and vice versa [63] . There were no significant differences in crossover rates between these gene categories ( p = 0 . 61 , bootstrap test ) , although the caste-biased genes had higher than average crossover rates ( 18% increase; p<0 . 01; average for coding regions = 240 ρ/kb ) . We next compared crossover rates in genes with biased expression in drones compared to workers and vice versa [64] . Here we found highly elevated recombination rates in worker-biased genes ( 50% increase compared to average; p<0 . 01 ) and decreased recombination rates in drone-biased genes ( 28% decrease; p<0 . 01 ) and unbiased genes ( 23% decrease; p<0 . 01 ) . We conclude that worker-biased genes have higher recombination rates compared to drone-biased genes , but not compared to queen-biased genes . These results suggest that genes with elevated expression in both female castes ( queens and workers ) tend to have higher recombination rates , rather than those specifically expressed in workers . We used two measures to estimate the potential association between levels of germline methylation and rates of crossing over in genes: 1 ) levels of CpGO/E in coding sequences and 2 ) estimates based on direct detection of methylated CpG sites in sperm and egg using bisulphite sequencing [62] . Genes were classified as HCpG and LCpG based high or low values of CpGO/E as described earlier . These two measures are highly correlated . We detect significant methylation in 39% of the LCpG genes and 14% of the HCpG genes . Out of all genes where we detect methylation , 60% of the CpGs are methylated in the coding sequence of LCpG genes compared to only 18% in the HCpG genes ( S6 Fig ) . We classified genes as HMET , LMET or UNMET based on the observation of high , low or undetected levels of methylation in the germline . The HMET category had significantly lower average CpGO/E compared to other categories . However , the UNMET class has a bimodal distribution of CpGO/E , where 33% of genes have values of CpGO/E <0 . 7 , which could potentially represent germline-methylated genes that were not detected experimentally . The average crossover rate among LCpG genes is only 29% of the rate esimated in HCpG genes ( p<0 . 01 ) , consistent with an effect of germline methylation suppressing recombination , particularly in HCpG genes ( Fig 4B ) . Inferred levels of methylation are strongly correlated with patterns of gene expression: female-biased genes tend to be HCpG and highly recombining , whereas male-biased genes tend to be LCpG and have lower recombination rates . These patterns also correlate with levels of genetic variation: LCpG genes have on average 45% lower genetic diversity than HCpG genes [37] . The association between levels of recombination and experimentally inferred levels of germline methylation is consistent with these results . Highly methylated genes have low levels of crossing over , similar to those observed in LCpG genes . A potential concern is that estimates of ρ made by LDHAT are affected by local variation in NE across the genome , which could lead to underestimation of recombination rate in regions of low genetic variation . Since ρ and θ are correlated , we conducted additional high resolution scans to test whether the differences in ρ we observe in coding relative to intergenic regions and in LCpG genes relative to HCpG genes could be due biases in inference caused by differences in local genetic diversity between these regions [37] . We measured ρ and θ in 1 kb windows across the genome . We found that ρ is consistently higher outside of genes than inside of genes at given levels of θ ( S7 Fig ) . Likewise , HCpG coding sequences are typically associated with higher ρ than LCpG coding sequences at given levels of θ ( S7 Fig ) , although the difference is less clear in regions where diversity is very high . We conclude that our inference of ρ in these regions detect significantly different crossover rates that are not merely mirroring local levels of genetic diversity . We next aimed to test whether the associations between caste biased gene expression , CpG levels , and crossover rates were indicative of specific gene categories being preferentially located in regions of certain recombination rates , or whether the association was restricted to recombination in coding sequences . Such a regional effect would be predicted if there was a selective advantage for worker-biased genes to occur in regions of high recombination [39] . We therefore compared patterns of gene expression and methylation to crossover rates in gene-flanking sequence , using 50 kb regions located 10–60 kb from each side of the genes ( Fig 4C ) and in 100 kb regions located 50–150 kb from each side of the genes ( Fig 4D ) . As expected , crossover rates increase with increasing distance from the gene ( average rate at 10–60 kb distance = 360 ρ/kb; average rate at 50–150 kb distance = 399 ρ/kb ) . In addition the associations between expression patterns and CpG levels are greatly reduced . The average decrease in crossover rates of drone-biased genes in flanking regions at 10 kb distance is only 13% compared to all genes and 4% of average >50 kb away . Crossover rates in the queen vs . worker comparisons are indistinguishable from the average rates in both >10 kb and >50 kb . The differences in crossover rates between LCpG and HCpG genes are also reduced in flanking regions compared to crossover rates within coding sequence . There is an 3 . 37x difference in crossover rates between LCpG and HCpG within coding sequence ( p<0 . 01 ) but this is reduced to 1 . 24x and 1 . 07x in the >10 kb and >50 kb flanking regions respectively . Associations between methylation classes and crossover rate are also significantly weakened in flanking sequence . These results indicate that crossover rates vary greatly between genes and correlate with both patterns of gene expression and levels of germline methylation . However , the finding that these associations are restricted to crossover rates in coding regions is indicative of a direct effect of these factors rather than an accumulation of certain types of genes in regions of high or low recombination , which would be predicted if there was an evolutionary advantage of worker genes being located in regions of high recombination [39] . We tested whether the associations between crossover rates and gene expression and CpG content were independent of each other . Genes that are biased in workers compared to drones are enriched in the HCpG class , so it is not clear which of these two factors is driving the association with high recombination rates . We therefore subdivided both datasets of caste-biased genes according to HCpG and LCpG classes . We found that the large differences in crossover rates in HCpG and LCpG remain irrespective of patterns of gene expression ( S8 Fig ) : for the same gene expression class , crossover rates are 2 . 3–3 . 4x higher in HCpG compared to LCpG genes . However , within each CpG class , the difference in crossover rates between drone and worker biased genes is smaller ( 1 . 3x higher in HCpG genes and 1 . 8x higher in LCpG genes ) . Hence , variation in CpG content is the strongest predictor of recombination rate in our dataset . One interpretation for this finding is that variation in levels of germline methylation is the strongest factor determining variation in recombination rates within genes in the honeybee genome . However , the associations we observe with crossover rates and gene expression patterns cannot be completely explained as an effect of differences in inferred levels of germline methylation . The site frequency spectrum in our dataset is dominated by low frequency AT alleles , which make up 80% of the rare variants ( allele frequency<10% ) across all SNPs , but only 51% of common variants ( allele frequency 40–50% ) , a highly significant difference ( p<10-5 , Fisher’s exact test; S9 Fig ) . By comparing homologous genomic regions between A . mellifera and A . cerana , we were able to infer the probabilities that either allele represented the ancestral or derived state at 2 , 983 , 700 SNPs using a weighted parsimony method ( see Methods ) . We categorised each allele at a SNP as weak ( A or T ) or strong ( G or C ) . At strong-to-weak ( SW ) SNPs , the S allele is ancestral and the W allele is derived , whereas weak-to-strong ( WS ) SNPs are defined as the reverse . The derived allele frequency spectrum consists mostly of strong-to-weak ( SW ) mutations ( 2 , 037 , 148 SNPs ) , and these are strongly biased towards occurring at low frequencies ( Fig 5A ) . Weak-to-strong ( WS ) mutations are fewer overall ( 719 , 365 SNPs ) , but are shifted toward high frequency or nearly fixed . Analysis of the proportions of variants of each type across the allele frequency spectrum therefore reveals a decline in SW and increase in WS variants with increasing allele frequency . WS variants make up 15% of variants at allele frequencies <0 . 1 but 79% of variants at allele frequencies >0 . 9 ( Fig 5B ) . This highly skewed site frequency spectrum is indicative of a strongly AT-biased pattern of mutation coupled with a fixation bias towards WS mutations . Such a fixation bias could be generated by a strong effect of GC-biased gene conversion ( gBGC ) , which manifests as a bias towards transmission of GC alleles . In order to further investigate this process , we quantified the average allele frequencies of a variety of classes of variants . We found that WS transitions segregate on average at 3 . 6x higher allele frequency in the population than SW transitions ( p<0 . 01; Fig 5C ) . However , the average frequencies of WS and SW transversions were similar to each other ( 14 . 1% and 14 . 9% , respectively ) and close to the average derived allele frequency in the sample ( 16 . 8% ) . These results are consistent with a fixation bias driven by WS transitions ( A→G or T→C ) , which could indicate that gBGC specifically targets transitions in the honeybee genome . A potential mechanism for this could be that heteroduplex mismatches between two alleles formed by a transition are repaired with a greater GC-bias than other mismatches in honeybees during meiosis . To our knowledge , such a mechanism has not been observed in any other species . We next tested whether gBGC could potentially be responsible for the huge excess of CpG dinucleotides observed in the honeybee genome . CpG sites are highly enriched in the genome ( CpGO/E = 1 . 64 ) but GpC occur at numbers close to the expected ( GpCO/E = 1 . 03 ) . This suggests an excess number of WS mutations that generate CpG sites occur or that they have a greater chance of fixation . We detect significantly elevated average frequencies of CpG-generating WS variants ( 0 . 43 ) compared with GpC-generating WS variants ( 0 . 37 ) in the population , although there is no difference between CpG and GpC generating SS variants , which are not expected to be affected by gBGC ( Fig 5D ) . The proportion of WS variants that generate CpG sites compared with those that generate GpC sites is 1 . 17 at low derived allele frequencies ( <0 . 1 ) but 1 . 72 at high derived allele frequencies ( >0 . 9; p<10-5; S10 Fig ) . Conversely , the proportion of SW variants at ancestral CpG sites compared with those that are ancestrally GpC is 1 . 73 at low derived allele frequencies ( <0 . 1 ) and 1 . 30 at high derived allele frequencies ( >0 . 9; p<10-5 ) . Hence , there appears to be a fixation bias in favour of CpG-creating mutations and against CpG-destroying ones . These results could explain the excess of CpGs in the honeybee genome . This suggests that fixation bias due to gBGC displays neighbour-dependency in honeybees , which has not been reported for any other species . In addition to gBGC , it is also possible that WS CpG-creating mutations could be positively selected if CpG were selectively maintained . We sought to investigate the dependency of the fixation bias due to gBGC on GC content and recombination rate . WS variants occur at higher frequency on average than SW variants in all GC and recombination rate categories ( Fig 6A and 6B ) . The difference between these frequencies increases as a function of both of these variables . For example , the average frequency of SW variants is reduced by 56% in regions of high GC ( 0 . 50–0 . 55 ) compared to low GC ( 0 . 15–0 . 20 ) but the average frequency of WS variants is only reduced by 25% . This indicates that the site frequency spectrum is more skewed towards high frequency WS alleles in regions of high recombination and GC content . This trend suggests that the strength of gBGC is stronger in regions of high recombination and GC content . We estimated the strength of the transmission bias due to gBGC in the honeybee genome using the model of Glémin et al . [57] . This method estimates the population-scaled gBGC parameter B , which is equivalent to 3NEb , where b is the transmission bias in favour of GC alleles . This method allows taking into account both polarization errors of mutations , which can lead spurious or biased signature of gBGC , and demographic effects distorting site frequency spectra ( see Methods ) . The maximum-likelihood estimates of B reveal a strong influence of gBGC on the fixation process of alleles . Average B in the genome is 5 . 71 , which is 15 times higher than average levels of B estimated from the site frequency spectrum in human populations ( 0 . 38 ) [57] . Levels of B this high are only found in the most extreme regions of the human genome that likely correspond to recombination hotspots [57] . Estimates of B vary between transitions and transversions ( B in transitions 6 . 47; B in transversion 0 . 03 ) . These estimates are consistent with our earlier inference that the effects of gBGC are restricted to transitions in honeybee . We estimated B in subsets of the genome divided according to GC content and recombination rates ( Fig 5C and 5D; S1 and S2 Tables ) . The association with GC content is strongest , and B increases from around 2 in the lowest GC content category ( <0 . 2 ) to a maximum of >15 in GC content >0 . 55 . The estimates of B increase from <1 in in the regions of lowest recombination to a maximum of around 7 in regions with crossover rates >400 ρ/kb . Even the lowest values of B are several times higher than the average in humans and some parts of the honeybee genome have extreme values of B . We expected B to be correlated with crossover rate , as gBGC is a recombination-associated process . However , here we find a stronger correlation with GC content . One reason for this could be that GC content is a more accurate indicator of recombination rates than our LD-based map because it is the result of the action of gBGC over evolutionary time . Another possibility is that our LD-based map predominantly measures crossover rates , which may not be strongly correlated with non-crossover rates . It is also possible that gBGC has a stronger correlation with non-crossover rates than crossover rates , as observed in humans [47] . The method also allows us to estimate the AT mutational bias , λ . We estimate the average bias over the whole dataset as 11 . 69 . The strong AT mutational bias appeared specific to transitions ( λ in transitions 13 . 09; λ in transversion 3 . 03 ) . Estimates of λ vary slightly across the genome . They are inferred to be higher ( 9 . 71–12 . 40 ) in regions of lower recombination ( <300 ρ/kb ) and lower ( 7 . 91—9 . 94 ) in regions of higher recombination ( >300 ρ/kb; S2 Table ) . When fitting the population genetics model we estimated high polarization error probabilities for WS mutations ( between 10 and 20% ) . We therefore performed simulations to estimate the potential effects of this on our estimates of B . We find that the high levels of gBGC may explain the high SNP polarization errors and that our estimates of B are robust to these errors ( S1 Text ) . Moreover , the high and significant skewness of the folded GC spectra ( see S1 and S2 Tables ) , which are not sensitive to polarization errors , is congruent with a force pushing GC content far from the expected mutational equilibrium .
Here we used patterns of linkage disequilibrium in 30 diploid honeybee genomes to estimate variation in crossover rates across each chromosome . Our results are consistent with previous results suggesting that meiotic homologous recombination occurs at extreme rates in honeybees [2–4] . The landscape of recombination rate variation in the honeybee genome does not appear highly punctate as found in a wide rage of other species [7–10] . We find a strong correlation between genetic variation and crossover rates indicative of pervasive influence of linked selection . Our results are also informative about the structure and organisation of the genome in relation to intensity of recombination , and identify specific factors that are likely to mediate recombination rates . We also show that recombination has extensive influence over population genetics and genome evolution in honeybees via the process of GC-biased gene conversion ( gBGC ) , which results in a bias in favour of fixation of WS mutations [42] . The strength of the bias in honeybees is an order of magnitude higher than previously observed in other species . Although our estimate of average crossover rate of 26 cM/Mb is similar to previous estimates , the correlation with a previous genetic map [3] is only moderate . It is possible that differences in the genome assemblies used by the studies contributed to these differences—our study used Amel_4 . 5 , whereas the Solignac et al . [3] study used Amel_4 . 0 . It is also possible that additional factors that affect patterns of LD , such as selection and gBGC , could affect our estimates of recombination rates . Temporal variation in recombination rates could also explain the moderate correlation between the two maps . The map by Solignac et al . [3] is based on markers segregating in the progeny of two queens thus corresponding to present-day recombination rates , whereas our map integrates all recombination events over a historical period . Another possibility is that there is variation between individuals in recombination landscape . In particular , we have focussed on samples from African honeybee subspecies , whereas the map by Solignac et al . [3] used two queens of European origin . In the future it would be interesting to investigate the genetic determinants of any inter-individual variation in recombination rates . Studies of fine-scale variation in recombination rate have revealed large variation in rates across the genome in a wide range of sexual eukaryotes [7–10] . These include plants [8] , fungi [46] , and vertebrates [7] . Conversely , invertebrates such as the nematode worm Caenorhabditis elegans [19] and the fruit fly D . melanogaster do not seem to have strong hotspots [20–23] . Interestingly , recombination does not seem to be required for synapsis in these species [65 , 66] , and mechanisms that are not dependent on sequence features may govern location of crossover events . The distribution of recombination events in honeybees also seems to follow this pattern . This suggests that a PRDM9-like protein that targets specific motifs during initiation of recombination is not present in honeybees further supports the notion that PRDM9 is a derived state in vertebrates [8] . The reasons for extremely high recombination rates in honeybees and other social insects are elusive . One possibility is high recombination rates are connected to the evolution of worker phenotypes , because the evolution of sociality specifically involved positive selection on worker behaviour [39] . This could potentially favour increased recombination rate in the vicinity genes involved in worker phenotypes because selection is more efficient in regions of high recombination [67] . Alternatively , high variability in worker phenotypes could be needed to maintain a stable and diverse workforce , which could also potentially favour increased recombination rates in the vicinity of genes involved in worker phenotypes [40 , 41] . A possible prediction of both of these scenarios is that genes with biased expression in the worker caste are preferentially located in highly recombining regions of the genome . Previous analyses of honeybee recombination found increased levels in worker genes , consistent with the above hypotheses [4 , 39] . However , here we report that a ) elevated recombination rates are observed in genes with biased expression in either of the female castes and not specifically in worker-biased genes , b ) pattern of gene expression are not well correlated with recombination rates in noncoding flanking regions , and c ) germline methylation patterns inferred by CpGO/E are more strongly associated with recombination rates than gene expression patterns are . Our data are consistent with a model where germline gene body methylation is the main modulator of recombination rate in genes and that correlations with gene expression are a side effect of this . Although evolution of eusociality likely involved strong selection for high recombination rates there is no evidence so far to indicate that it involved increases in recombination rate in specific genomic regions . Both housekeeping genes , and genes mainly expressed in drones , are inferred to be germline methylated and have suppressed levels of crossing over [37 , 61] . Genes with high CpGO/E have crossover rates similar to intergenic regions . These observations are consistent with the view that DNA methylation is the main cause of reduction of recombination rates in genes and variation in recombination rates between genes , although we cannot rule out the effect of another factor indirectly associated with methylation . It is important to note that the association between recombination and CpGO/E in the honeybee genes could also be influenced by gBGC , which generates new CpG sites [68] . However , the link between CpGO/E and methylation in honeybee genes is well established [60 , 61] and confirmed in this study . Methylation is generally restricted to gene bodies in honeybees ( 75% of methylated CpGs are found in exons [60] ) and recombination rates in noncoding regions are higher outside of genes . Methylation has been suggested to suppress recombination rate in a variety of species including the fungus Ascobolus [69] and angiosperm Arabidopsis [70] . Vertebrate genomes tend to be highly methylated , but hypomethylated CpG islands have elevated recombination rates in some species [18 , 71 , 72] . We therefore hypothesise that germline DNA methylation suppresses recombination in honeybees . We find a strong correlation between recombination rate and levels of neutral genetic variation that remains after correcting for mutation rate inferred from levels of divergence with an outgroup , A . cerana . Similar correlations are observed in a diverse range of species and are believed to reflect the effects of recurrent selective sweeps ( positive selection ) and/or the effects of selection removing linked deleterious variants ( background selection ) [28 , 32] . If selection occurs at similar rates across the genome , then it will have a greater effect on linked variation in regions of low recombination leading to this general correlation . The predicted effect of selection depends on the rate at which it occurs across the genome and whether variants are strongly or weakly selected . Interestingly , we also find that the correlation between diversity and crossover rates is weaker in regions far from genes , which is consistent with a lower density of functional sites and hence potential targets of selection in these regions . Our findings are therefore consistent with a pervasive impact of selection on genome variation in honeybees similar to inferences in fruit flies [32] . Recombination increases the efficacy of selection [29 , 67 , 73] and high levels of selection ( e . g . due to recurrent selective sweeps ) are a potential explanation for the extreme recombination rates observed in honeybees . Our analysis also indicates a dominant effect of gBGC on genetic variation in honeybees . The derived allele frequency spectrum contains a large excess of SW mutations segregating at low frequencies , and an excess of WS mutations at high frequencies close to fixation . This skewed site frequency spectrum is indicative of a strongly AT-biased pattern of mutation and a fixation bias towards WS mutations , consistent with a strong effect of gBGC . However , our analysis indicates that gBGC in the honeybee has two features that have not been reported in other species . First , the WS fixation bias appears much stronger for transition than transversion mutations , which could reflect a greater strength of GC-bias in repair of mismatches caused generated by transitions during recombination . Second , we find evidence that this bias is stronger in CpG compared to GpC sites . This suggests that the repair bias could also be neighbour dependent in honeybees . This could explain the massive excess of CpG sites observed in the honeybee genome . The reasons for these specific biases are unclear . Quantification of gBGC in humans found no evidence for repair bias towards transitions or CpG sites [57] and no such biases have not been observed in other species either . In vivo experiments in mitotic mammalian cells suggest that G/T mispairs in DNA , which can be generated by transitions , are strongly biased towards being repaired to GC rather than AT [74 , 75] . However , these biases result from the base excision repair ( BER ) pathway , and mismatches during recombination are mainly repaired by mismatch repair ( MMR ) . Our results could therefore indicate a greater role of BER in repair of mismatches during recombination in honeybees , or they could suggest that these biases occur in MMR in honeybees . It has been suggested that such repair biases exist in order to correct common types of mutations , in particular due to hypermutability at methylated CpG sites in mammals [76] . We observe a strong AT mutation bias , particularly in transitions , which is counteracted by a strong GC fixation bias in transitions . Levels of methylation and CpG mutation in honeybee are generally low , but are restricted to genes . Mutations at such sites could be more accurately repaired by the CpG-biased mechanisms we infer here . We estimate strength of the fixation bias due to gBGC in honeybees to be incredibly high ( average B = 5 . 71 ) , and much greater than observed previously in any other species ( average B in humans is 0 . 38 ) . Such a high level of gBGC is likely recent because the average GC content of the honeybee genome ( 0 . 34 ) is much lower than the equilibrium GC content predicted by the balance between gBGC and AT mutational bias ( GC* = 1/ ( 1+λe-B ) ≈ 0 . 96 ) . At values of B less than one , as observed in the human genome , gBGC is not expected to dominate over random genetic drift [57] . However the values of B estimated here are substantially greater than one , indicative of a dominant influence on molecular evolution . Indeed , across much of the genome , they exceed B = 8 . 7 , the value estimated for human hotspots , which is expected to result in the fixation of a substantially elevated number of deleterious nucleotide substitutions [54] . The magnitude of B depends on both effective population size , NE , and the transmission bias in favour of GC alleles , b ( B = 4NEb for diploids and B = 3NEb for haplodiploids ) . Using estimates of NE of 10 , 000 [77] for humans and 500 , 000 for honeybees [37] leads to estimates of b of 9 . 5 x 10-6 and 3 . 8 x 10-6 , respectively . Hence , we infer that the transmission bias in humans should be 2 . 5 times stronger than in bees . However , due to the higher NE in bees , this lower transmission bias still has an extreme effect on the allele frequency distribution . The honeybee is still unusual in having extremely high levels of gBGC , as related taxa with high NE do not seem to have similarly high levels . In particular , the site frequency spectrum in D . melanogaster does not appear strongly skewed [78] . It therefore seems likely that the extreme recombination rates in honeybees are linked to high levels of gBGC , even if the transmission bias in meiosis is not greater in magnitude than humans . In addition , compared to Drosophila , it is also possible that the high AT mutation bias in honeybees has selected for a stronger b per meiosis . It should also be noted that recombination only occurs in honeybee females , which suggests that the transmission bias in female meiosis is likely to be twice our estimate here , which is a sex-averaged estimate . The strong skews in site frequency spectrum and fixation biases are incompatible with a standard model of population genetics whereby the fate of alleles is determined by genetic drift and selection . The process of gBGC has a major influence on probability of fixation of an allele in honeybee populations . This has major implications for molecular evolution , as it can interfere with the removal of harmful alleles and fixation of beneficial alleles by natural selection and cause fixation of weakly deleterious mutations . Selection for higher recombination rates in honeybees therefore appears to have entailed the considerable additional cost of strong gBGC .
We aimed to produce a high-resolution map of recombination in the Western Honeybee Apis mellifera using 30 diploid sequences from African worker bees collected in South Africa and Nigeria . Although these populations are geographically separated , analyses of population structure suggest that this sample can be regarded as panmictic and a single population . The bees were sequenced as part of a different study and short read mapping , genotype calling , filtering and phasing procedures are described in Wallberg et al . [37] . Watterson’s estimator [79] was used to calculate the population mutation rate per base ( θw ) as a measure of genetic diversity across the genome . Diversity , GC content and CpGO/E was calculated in windows of 100 kb along the chromosomes using the current reference genome ( Amel_4 . 5; [80] ) . These statistics were averaged across the full window and for each type of functional element ( coding , intron , UTRs and intergenic sequence; coordinates according to the recent gene annotations in OGSv3 . 2; [80] ) in the window . The African population includes 6 . 2 million single nucleotide polymorphisms ( SNPs ) , corresponding to an average level of genetic diversity of θw = 0 . 008 . The reversible-jump MCMC algorithm interval of the LDHAT program [81] was used to estimate the mean population-scaled recombination rate coefficient ρ ( rho ) across regions ( or intervals ) , which in honeybees is taken as ρ = 3NEr ( 3NE is due to honeybee haplodiploidy ) and where r is the genetic map distance over a region . The interval method fits a uniform recombination rate over a region from patterns of linkage disequilibrium ( LD ) among genotypes . The LDHAT recombination map ( hereafter referred to as the LD map ) was estimated along the chromosomes in segments of up to 2 , 000 variable sites . The segments were arranged to never span across scaffolds and had an average physical length of 63 kb . For each segment , the interval program was run for 1 . 1 million iterations and the chain was sampled every 10 , 000 iterations , following a burn-in of 100 , 000 iterations . We evaluated the performance of different block penalties ( see below ) . A map inferred with a block penalty of 1 was taken as the canonical LD map for the study . Levels of genetic diversity are highly variable along the honeybee chromosomes and correlates with functional elements caste biased expression and nucleotide composition [37] . We therefore performed an analysis to determine whether our method could be biased towards detecting high recombination in regions of high SNP density . Three measures were put in place in order to study the effect of local diversity and LD on the inference of broad-scale recombination from our data: The LDHAT recombination map ( hereafter referred to as the LD map ) was compared to the GC and CpGO/E composition computed across the full length of each gene in the OGSv3 . 2 gene annotation and according to intervals of each type of functional element . Gene lists with accessions associated with biased gene expression between queens and workers [63] , as well as between drones and workers [64] were queried in order to further assess the interaction between recombination and caste function . The gene lists were subdivided into classes of low or high CpGO/E in order to facilitate analyses of the influence of both sequence composition and caste function on recombination . The significance of differences in crossover rates between gene expression and low or high CpGO/E categories were measured using a bootstrap test . We randomly resampled 200 pseudo-replicates from each class and compared their values in order to generate confidence intervals and estimate significance . We estimated levels of germline methylation in genes using data from Drewell et al . [62] . Significantly methylated CpGs ( mCpGs ) were originally detected using short read bisulfite sequencing of honeybee egg and sperm cells and mapped against v2 . 0 of the honeybee genome . In order to estimate methylation levels in different genes , we merged the two methylation tracks into a single germ line track and associated the coordinates of the mCpGs with overlapping coding sequences using the matching gene model annotation ( OGSv1 . 1 ) . We next measured methylation levels in two ways for each accession: i ) the number of mCpGs per kb of coding sequence ( controlling for the length of the gene ) ; and ii ) the proportion of CpGs in the coding sequence of a gene that were methylated ( controlling for the actual CpGs available to methylate ) . We then used BLAST to link OGSv1 . 1 accessions to the current OGSv3 . 2 accessions , for which we have estimated CpGO/E and recombination rates . 8901 genes were linked across the two annotation systems and included in the downstream analyses . Out of the 8901 genes , the coding sequence of 2449 genes were found to be methylated in at least one CpG site whereas 6452 genes had no evidence of methylation and were classified as unmethylated ( UNMET ) . We divided the methylated genes into two equally sized low methylation frequency ( LMET ) and high methylation frequency classes ( HMET ) . We estimated the average crossover rates for these categories ( UNMET , LMET , HMET ) and generated 95% confidence intervals from 200 bootstrap replicates . The LD map was compared to an experimental recombination map ( hereafter referred to as the genetic map ) produced by Solignac et al . [3] from parent-offspring recombinant frequencies inferred from >2 , 000 evenly spaced microsatellite markers . The markers and genetic distances of the genetic map had originally been computed for an older version of the genome ( Amel_4 . 0; 183 Mb ) . In order to facilitate a 1:1 comparison between the two methods , we identified the locations of the corresponding marker coordinates for Amel_4 . 5 ( 229 Mb ) using BLAST [82] of 2 kb flanking sequence associated with each marker . Out of the 2008 original markers , 1974 markers could be mapped unambiguously to Amel_4 . 5 . The remaining markers were not included due to primer sequences aligning to different scaffolds or chromosomes or at unexpectedly large distances from each other compared to the original positions . Between the two versions of the reference genome , there had been extensive reorganisation and reorientation of scaffolds . Many genetic distances had originally been estimated across scaffolds , which themselves may have been subject to change . By querying multiple 2 kb segments of each of the v4 scaffolds against the v4 . 5 chromosomes with BLAST , we detected orientation changes in 124 out of 371 scaffolds ( 33% ) . These changes often caused previously adjacent markers to be separated by additional markers on the new reference sequence , resulting in overlapping genetic intervals and a much-reduced average recombination rate of 11 . 3 cM/Mb across the genome , compared to the reported rate of 22 cM/Mb . By including genetic distances stretching across adjacent scaffolds only if they were both plus-oriented , we produced a new genetic map with an average rate to 22 . 3 cM/Mb , which was next correlated to the LD map in windows of 1 Mb . The last window of each chromosome was only included if it spanned at least 0 . 5 Mb of sequence . The Eastern honeybee A . cerana is a sister species of A . mellifera and was used as outgroup in several analyses . Short reads from 10 diploid worker samples were mapped as described in ref . [37] and pooled in order to produce an A . cerana consensus sequence from sites with a minimum depth of coverage of 5x . The consensus sequence was next used to estimate the nucleotide divergence between the two species and use the outgroup allele to infer the ancestral state at A . mellifera SNPs . At sites where the ingroup is polymorphic ( X|Y ) and the outgroup is fixed for one of the two alleles ( e . g . X ) , simple parsimony assumes that the allele shared between the ingroup and the outgroup is the ancestral allele ( X ) and that a X→Y mutation generated the polymorphism in the ingroup . However , this reconstruction does not take into account the possibility that the other allele ( Y ) was the true ancestral allele but was substituted in one of the species ( Y→X ) , followed by an Y→X mutation which generated the X|Y polymorphism in the ingroup . To incorporate this uncertainty and reduce the error in the polarization of the mutations , we applied a weighted parsimony method that incorporates substitutions to estimate the conditional probabilities that either allele represent the ancestral or derived state given an ingroup polymorphism and an outgroup allele [49] . The polymorphisms were next classified as transitions ( Ti ) or transversions ( Tv ) and whether they were weak-to-strong ( WS ) , strong-to-weak ( SW ) ; weak-to-weak ( WW ) or strong-to-strong ( SS ) , whereby a weak allele is A or T and a strong allele is G or C . In total , 3 . 02M SNPs were classified according to this scheme . The average population frequency of the derived allele ( f D ) of each SNP was estimated across the genome and related to dinucleotide context , regional GC content and recombination rates ( computed from windows of 100 kb ) . We used the method of Glémin et al . [57] to estimate the strength of gBGC . In brief , this method fits a population genetics model to the derived allele frequency ( DAF ) spectra of the three kinds of mutations , 1 ) W→S , 2 ) S→W , and 3 ) S→S and W→W . This model takes into account the departures from the equilibrium induced by demography , population structure and/or sampling . Despite modelling an explicit demographic scenario , the model includes fuzzy parameters correcting for the distortion of the spectrum compared to the one expected in an equilibrium population , following the approach of Eyre-Walker et al . [83] and as initially implemented for gBGC in Muyle et al . [84] . Importantly , it also corrects for polarization errors of mutations that can bias gBGC estimates [85] . Because , it was proved to be difficult to estimate the heterogeneity of B without additional information to constrain the model [57] , we only fitted a constant gBGC model ( model M1* in [57] ) . Given the average GC-content , the AT mutational bias can also be estimated . To get the DAF spectra , the numbers of SNPs detected in each mutational class were summed over the frequency spectrum across the whole dataset . Site frequency spectra were also generated according to bins of local GC ( 100bp window to either side of the SNP ) and regional recombination ( 1000bp window ) . We estimated B for all mutations and for transitions and transversion separately . | Evolution results from changes in allele frequencies in populations . The main forces that cause such changes are natural selection and random genetic drift . However , an additional process , GC-biased gene conversion ( gBGC ) , associated with meiotic recombination , affects the probability that alleles are passed from one generation to the next . The honeybee , Apis mellifera , has extremely high recombination rates—more than 20 times to those observed in humans . However , the reason for this is unknown and the effects of such high recombination rates on evolution are not well understood . Here we use patterns of genetic variation in the genomes of 30 honeybees to infer variation in the rate of recombination across the genome . We find that recombination rates and levels of genetic variation are strongly correlated , which is indicative of a pervasive impact of natural selection on genetic variation . We also infer a major role of DNA methylation in determining recombination rates in genes . Patterns of genetic variation appear to be strongly skewed due to the effects of gBGC , suggesting that recombination generates a bias in transmission of alleles during meiosis . This process seems to be interfering with the efficacy of selection at removing deleterious alleles and favouring beneficial ones . Recombination therefore has a huge impact on genetic variation and evolution in honeybees and appears to play a dominant role in genome evolution . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [] | 2015 | Extreme Recombination Frequencies Shape Genome Variation and Evolution in the Honeybee, Apis mellifera |
Rates of hospital-acquired infections , such as methicillin-resistant Staphylococcus aureus ( MRSA ) , are increasingly used as quality indicators for hospital hygiene . Alternatively , these rates may vary between hospitals , because hospitals differ in admission and referral of potentially colonized patients . We assessed if different referral patterns between hospitals in health care networks can influence rates of hospital-acquired infections like MRSA . We used the Dutch medical registration of 2004 to measure the connectedness between hospitals . This allowed us to reconstruct the network of hospitals in the Netherlands . We used mathematical models to assess the effect of different patient referral patterns on the potential spread of hospital-acquired infections between hospitals , and between categories of hospitals ( University medical centers , top clinical hospitals and general hospitals ) . University hospitals have a higher number of shared patients than teaching or general hospitals , and are therefore more likely to be among the first to receive colonized patients . Moreover , as the network is directional towards university hospitals , they have a higher prevalence , even when infection control measures are equally effective in all hospitals . Patient referral patterns have a profound effect on the spread of health care-associated infections like hospital-acquired MRSA . The MRSA prevalence therefore differs between hospitals with the position of each hospital within the health care network . Any comparison of MRSA rates between hospitals , as a benchmark for hospital hygiene , should therefore take the position of a hospital within the network into account .
Pathogens that typically cause hospital-acquired infections have an opportunistic nature . These organisms are usually part of the normal bacterial flora of humans and only cause disease when reaching body sites that are normally free from bacterial colonization e . g . when anatomical barriers are breached due to trauma or medical/surgical interventions . For this reason , severe problems with nosocomial pathogens are mainly seen in the very young and elderly and most frequently in institutions such as hospitals and long-term care facilities where patients are treated for acute or chronic conditions . Methicillin-resistant Staphylococcus aureus ( MRSA ) is an antimicrobial resistant variant of S . aureus , a common bacteria frequently colonizing healthy humans and animals . Emergence of MRSA is due to the acquisition of a large DNA fragment , which seems to be rare [1] , [2] . The expansion of a limited number of MRSA clones that characterizes the current epidemic in hospitals worldwide is therefore believed to be the result of between patient transmission and only to a minor extent due to the ‘de novo’ emergence in patients exposed to antibiotics . MRSA has therefore become the marker with which the success or failure of hospital infection control [3] . The prevalence of the MRSA differs considerably within and between countries [4] , [5] . Currently about 30% of the S . aureus causing bloodstream infections in the UK is resistant to methicillin , against only 1% in the Netherlands and Scandinavian countries [6] . Although in high endemic countries MRSA infections are frequent in all hospitals , the proportions are highest in large teaching ( tertiary care ) hospitals [4] , [7] , which also report the highest frequency of newly occurring MRSA clones [8]–[11] . The severity of underlying medical condition of the patients , as well as higher antibiotic use and frequency of invasive procedures have been proposed as the main reasons for this difference [3] . Patients can carry MRSA , asymptomatically , for a long time [12] . When readmitted , they may introduce the pathogen acquired during a previous admission into a new hospital [13] . Failure of one hospital's infection control measures can therefore affect the prevalence in hospitals with which it shares patients [14] . Patients are referred to hospitals at different rates depending on the function of hospitals within the health-care system , which likely affect the prevalence at different institutions . These referral patterns might therefore offer an explanation for high MRSA incidence in hospitals of the tertiary referral level [7] . But can referral patterns account for differences in spread between hospitals , and for differences in observed prevalence ? To answer these questions , we have been mapping the health care network based on a large national medical registry , and evaluated the occurrence of hospital-acquired infections in different care categories under simulated epidemic conditions .
In 2004 , hospital care in the Netherlands was provided through 71 general hospitals , 19 top clinical hospitals and 8 university medical centres ( Figure 1A ) . During the observation period of one year ( 2004 ) 1 , 676 , 704 patients were admitted from the population of 16 . 7 million . These patients were admitted for a total of 2 , 611 , 452 times , the majority of patients were hospitalised once . The frequency with which patients were readmitted showed a right-skewed distribution ( Figure 1B ) , with still 86 patients being readmitted for more than 52 times . Patients stayed on average 4 . 3 days per hospital admission , patients who had less hospital admissions stayed longer per admission ( Figure 1C ) , and those who had four hospital admissions had on average the longest ( 5 . 6 days ) episodes of hospital admission . Moreover , these patients had the highest rate of readmission in different hospitals ( Figure 1E&F ) , whereas patients who were readmitted more frequently tended to return to the same hospital . These frequent attendees were also most likely to stay for only one day . The individual-based model emulated the dynamics of patient referrals and allows us to assess the spread of hospital-acquired infections . Colonized patients from one hospital spread the pathogen to nearby hospitals within days , but it takes more time –5 to 10 years– before all hospitals encounter it ( Figure 2A ) . The median time to first infection ( TFI ) for university medical centers ( UMCs ) was 755 days , the TFI for top clinical hospitals was 1 , 087 days and the TFI for general hospitals was 1 , 346 days . At any stage of the epidemic the expected prevalence in UMCs was higher than in general and top clinical hospitals ( Figure 2C ) . We reconstructed the Dutch national network of hospitals ( Figure 3A ) with respect to the potential spread of hospital-acquired infections , using patient referral patterns taken from national medical registration ( LMR [15] ) . Within this network , the UMCs show a higher degree of connectedness than the general and top clinical hospitals ( Figure 3B ) . General hospitals had a higher outdegree than indegree , whereas the reverse was true for UMCs , resulting in an 8-fold difference in the indegree between both types of institutions . Top clinical hospitals assumed an intermediate position and showed little difference between indegree and outdegree . Moreover , the indegree relative to the total number of admissions ( including patients admitted directly from the community ) was much higher in the UMCs compared to the general hospitals . The patient flow through the network was thus directed towards the UMCs . In order to determine the effect of the directionality of the network , we repeated the analysis of the individual-based model using a dataset with alternative direction . We created a dataset in which all referral probabilities to hospitals were set equal . In the resulting network , both the indegree and outdegree of the UMCs were higher than the other hospital categories , but the outdegree is now higher than the indegree ( Figure 4A , B & C ) . The relative indegree was higher for the general hospitals compared to the other two categories , although there was only a small difference between UMCs and top clinical hospitals . These simulations resulted in slightly higher prevalence in the general hospitals , compared to the top clinical hospitals and UMCs . The differences between the hospitals in connectedness and prevalence are caused by the different hospital sizes , the only parameter that varied between hospitals in this model . This suggests that the short time to first infection of UMCs is due to their absolute high degree of connectedness , while their high relative indegree causes the higher prevalence in UMCs relative to other hospital categories . We also used two other networks with alternative directions , to test if our observation holds under different conditions . First , we reversed the direction of the network by reversing time in the original dataset , the patients who first visited a general hospital and then a UMC now do the opposite . In this dataset the UMCs still have a higher relative indegree , compared to the general hospitals , although their outdegree is now higher than their indegree ( Figure 4D , E & F ) . These simulations reduced the difference in prevalence between hospitals , with still the highest prevalence in the UMCs . This exact reversion had almost no effect on the TFI of all hospital categories . Second , we increased the reversed direction in order to decrease the relative indegree of the UMCs to a level below the relative indegree of the general hospital , while keeping both the absolute degree of the UMCs ( both indegree and outdegree ) above the degree of the general hospitals . These simulations resulted in a lower prevalence in the university medical centres compared to the hospitals of other care categories , whereby the top clinical hospitals had the highest prevalence , reflecting their highest relative indegree ( Figure 4G , H & I ) . This reversion of direction in the network had , just like the previous ones , little effect on the order of TFI for the hospital categories . The results of all three simulation studies with alternative directions , when taken together , strongly suggest that the high prevalence in UMCs relative to other hospital categories is due to directionality of referral patterns , reflected by their high relative indegree .
This study sets a precedent by using data about all hospital admissions obtained from the National Medical Register ( LMR [15] ) to explore the potential spread of hospital-acquired infections through the Dutch national network of hospitals and describing the effect of nationwide referral patterns on the spread of nosocomial infections like MRSA . This method shows properties of hospitals , such as connectedness within the network , that on the level of a single hospital would not be visible . In the Netherlands , 98 hospitals provide various forms of specialist care . Within the category of general hospitals , there are considerable differences from hospital to hospital , with some smaller hospitals providing only basic hospital care . Therefore , patients who need advanced medical treatment need to be referred to so-called top clinical hospitals or university medical centres . Top clinical hospitals are large institutions that provide a wide range of clinical specialities and are involved in specialists training and education of doctors and other health care workers . In contrast to university medial centres they are not affiliated with universities and do not include the same comprehensive spectrum of specialities . Within the health care system , the university medical centres occupy a special place as leading hospitals with advanced specialist and final referral functions . In the Netherlands the hospital admission rate is rather low compared to international standards with 15 . 6 admissions per 100 inhabitants [16] and an average stay of only 4 . 3 days . This figure is low , as it also includes day care treatment when patients occupied a bed but do not stay overnight . The majority of patients ( 73% ) are admitted only once to any hospital . Few return twice ( 17% ) , three times ( 5% ) , or more ( 5% ) . Importantly , patients who are admitted twice or three times in a one year period not only have the longest per admission treatment episodes , but are also more frequently readmitted to different hospitals . In this way , all hospitals in the Netherlands become connected and form a network consisting of referred patients who form a bridge between hospitals and provide a path that can facilitate the spread of hospital-acquired infections , such as MRSA , between hospitals . The individual-based model which emulates the referral characteristics recorded in the LMR , describes the spread of nosocomial infections among hospitals on an individual patient level . It shows that patients who are admitted only two or three times contribute significantly to the inter-hospital spread of the infection and suggests that the prevalence is directly related to the referral level of different hospital categories . This model is , however , unable to provide a mechanistical explanation for the predicted differences in prevalence between hospital categories . For this reason , a simplified model of the hospital network was created . This model weights the contact pattern between hospitals on the basis of average patient referrals between any two hospitals without taking individual referrals and catchment populations into consideration . Despite being a coarse simplification , the hospital network model provides excellent heuristic value as it is able to demonstrate the directionality of the entire network , which is the driving force behind the difference in prevalence between different hospital categories . Our methods rely on three key assumptions that should be addressed . First , all of our methods do not take account of transmission outside of the hospitals . If community transmission of hospital-acquired infections become a significant factor , the dynamics of the epidemic will ultimately change and the effect of patient referrals between hospitals will be diluted . Community transmission of MRSA is mainly seen in families [17] , among military recruits [18] , in relation with competitive sport activities [19] and among children in day-care centres [20] . Typical community-acquired ( CA- ) MRSA is a phenomenon widely described in the USA [21]–[24] but still rather uncommon in Europe . Although CA-MRSA has been identified in Europe in countries with high as well as low MRSA prevalence , it so far remains much less prevalent than health-care associated ( HA- ) MRSA . Indeed a recent comprehensive study among patients consulting general practitioners in the Netherlands could not find any CA-MRSA in this population [25] . For MRSA , our models will lose validity when CA-MRSA becomes widespread in the general population and the prevalence in the population reaches levels comparable with those in hospitals . Second , we have assumed a specific measure of connectedness to create the network . However , the construction of hospital networks can be done based on other measures than the one we used , like weighting the contact between two hospitals by the number of patients these hospitals share , or by taking only subsequent admissions into account . These measures would slightly alter the difference in connectedness between the hospital types , but the differences between referral levels would remain ( data not shown ) . However , we feel that exclusion of data about the length of stay and time between admissions would disguise the true utilization patterns that govern the spread of HA-MRSA . Third , both the individual based model and the measure of connectedness assume homogeneous mixing within the hospital and leave out any ward structure . However , because the medical condition of a patient determines both the ward of admission and his/her health-care use , patients with a certain utilization pattern may mainly meet patients with comparable utilization patterns . This assortative behavior of patients [26] can potentially alter the dynamics of the epidemic , and especially the rate of growth of the epidemic . However , although the different wards may show different dynamics with the different patients they admit , the general direction of the referred patients will still be towards the university hospitals . We therefore expect the difference between hospital categories to still hold in the long run , despite some likely transient effects during the growth of the epidemic . A higher prevalence of health care-associated infections has been repeatedly demonstrated for tertiary referral centres such as university and teaching hospitals , which also witness the majority of outbreaks of these types of infections . As a conventional explanation , the severity of underlying conditions , more invasive diagnostic and therapeutic procedures and higher rates of antibiotic prescription have been incriminated for this difference . Our model predictions based on the observed admission pattern in the Netherlands , however , suggest a more parsimonious explanation . In the Dutch health care network , the university medical centres admit a large number of referred patients from other hospitals , much more than the top clinical hospitals ( Figure 3B ) . Each university medical centres is therefore connected to a large number of general hospitals as well as a number of top clinical hospitals . This central position within the hospital network puts these hospitals at higher risk of encountering colonized patients . Moreover , the flow of infectious patients through the hospital network is directed towards the university medical centres and we could show that as a direct result of this directionality , prevalence in these hospitals is predictably higher relative to the other categories . These observations can have important implications concerning hospital infection control . When hospital infection control fails within a single hospital , hospital-acquired infections will start to spread between hospitals , with the most connected ones at the highest risk of both acquiring and spreading the disease . Differentiation of intervention measures over hospital categories , for instance by making the university medical centres the focal point , could then be considered . The exact implementation of such a differentiation is , however , beyond the scope of this paper and should be the focus of further research . Furthermore , our results suggest that differences in prevalence of nosocomial infections between hospitals do not necessarily reflect the success of the hospital infection control measures of individual hospitals . Direct comparisons of infection rates between hospitals may therefore give a distorted view of hospital standards , if national ( or regional ) health care utilization patterns are not considered . The use of such comparisons , for benchmarking , may therefore lead to a false conclusion about a hospitals effort to reduce nosocomial infections . In summary we predict that ( 1 ) Hospital-acquired infections can spread rapidly from index hospitals to the next referral level . ( 2 ) Secondary and tertiary referral hospitals must be prepared for rapid response . ( 3 ) High connectedness and the directionality in the health care network towards the university medical centres cause a local build-up of nosocomial pathogens , such as MRSA , and thus a higher prevalence in these hospitals . This should be taken into consideration for benchmarking and the design of national control strategies .
We used the Dutch national medical register from 2004 ( Landelijke Medische Registratie LMR [15] ) , which contains the data about all individual hospital admissions for the total of Dutch hospital organizations of that year . We stratified patients in the LMR based on the number of admissions , , in the one year of data . Per stratum we counted the number of patients , , and measured the distribution of the length of stay , , the time between admissions , , number of hospitals visited , , and the changes between hospitals , . We defined a change between hospitals as an admission to a hospital different from the hospital of the previous admission . For each hospital we counted the number of next admissions in other hospitals to determine the referral probability , , and counted the the number of admissions per hospital to determined the size , . For reasons of privacy protection , we were not authorized to use the data at individual record level for detailed analysis . We therefore generated a simulated dataset based on the recorded frequencies which describes the individual patient referral patterns that is consistent with the observed patient characteristics in the LMR . This also enabled us to expand the simulated dataset beyond the recorded single year in the LMR to 20 years . We assumed that each patient's health-care use comes in sequences of a given number of hospital admissions , , and that the time between these sequences , i . e . between the moment of discharge of the last admission in the sequence and first admission in the next sequence , is exponentially distributed . Patients were assigned a hospital of initial admission from the hospital size distribution , , and a number of admissions in this sequence from distribution . The number of changes between hospitals during these admissions was picked from the distribution . If the number of changes was larger than 0 , the same was done for the number hospitals visited , picked from the distribution . We assumed that the moment of changing between hospitals was distributed uniformly over the admissions and the choice for the new hospital was based on the current hospital's referral distributions . The length of stay was picked from distribution and time between admissions from distribution for all sequential admissions . We picked the rate of initial admission , , based on over 1 . 6 million admitted patients for an entire population of 16 million individuals , at 1/3650 day−1 . After the last admission in the sequence , the time to next admission is therefore picked from an exponential distribution with mean . Because the average time between admission sequences is much longer than the average length of colonization , we thus assumed that the colonization status of an individual at the start of an admissions sequence does not depend on this individuals colonization status in the previous admission sequence . We created a dataset for 20 years to allow the epidemic to reach equilibrium level . Using the individual entries of the simulated dataset we subsequently created a mathematical model that describes the effect of individual patient movements through the hospital network on the spread of hospital-acquired infections . These individuals can either be susceptible or infected . No distinction was made between colonization and clinical infection for the sake of simplicity . Infected individuals ( ) infect susceptible individuals ( ) within the same hospital during one day with rate , where is the total number of patients in the hospital . Therefore , each susceptible has a probability of of getting infected per day . We assume that infectious patients spread the infection to a random sample of the patients within the hospital , and take no ward structure into account . Individuals lose the infection with rate and the mean duration of colonization was set at 365 days [12] . In order to explore the dynamics , we infect 10% of the patients that are admitted to an index hospital on a randomly chosen starting date , and monitor how the infection spreads to other hospitals . The number of colonized individuals at each time step and the time to first encounter of a colonized patient in each hospital ( time to first infection , TFI ) was recorded . For each index hospital we perform 200 simulations , sequentially repeating these sets of simulations for each 98 hospitals as index hospital , thus performing a total of 19600 simulations . In further analysis , we only include simulation runs resulting in an outbreak larger than a threshold of 1000 colonized persons , to exclude runs that resulted only in small local outbreaks . The results are not sensitive to the exact value of this threshold . In order to reduce the complexity inherent to the individual-based model , we created a hospital network model assuming transmission parameters between hospitals . All transmission parameters were based on the patient characteristics as observed in the LMR . Thereby , we calculated the infection rate , , from hospital to hospital , using the probability that any referred patient transmits the infection after referral . This probability depends on the patient's length of stay in both hospitals and the rate of losing colonisation between admissions . The infection rates between all hospitals form a 9898 matrix , , which describes the national network of hospitals in terms of potential transmission . For each admission we calculate the probability that the patient transmits the infection from the referring hospital to the admitting one , . This probability can basically be divided into three separate probabilities: contracting the infection in a referring hospital , , still being colonized on readmission , , and spreading the infection in the admitting hospital , : ( 1 ) The probability of being colonized depends on the length of stay in each referring hospital , , the number of colonized patients in each of these hospital , , and the transmissibility of the pathogen , ; . If we assume that both the infectivity and the number of colonized patients are at a fixed low level , we can simplify this to , where encompasses the transmissibility and low prevalence in the hospital . Because we assume the transmissibility and prevalence are equal in all hospitals , and because the matrix scales linearly with we can leave at unity: ( 2 ) The probability of introduction in the admitting hospital , , in turn depends on the length of stay in the admitting hospital , , the number of susceptible patients , , and the transmissibility of the pathogen , ; . Here , we can assume that the number of newly infected patients is not dependent on the size of the hospital , because ward size is generally not related to hospital size . Therefore , the probability of transmission is directly related to the basic reproduction number per admission , , and becomes . Where denotes the average length of stay in the dataset . Just as before , we assume that the number of colonized patients is low , and the process is not limited by the number of available susceptible individuals: ( 3 ) The probability that a patient is still colonized upon readmission , , depends on the time between discharge and admission , , and the recovery rate , ; . Although overlapping admissions do occur in the data –patients can for instance be moved to another hospital for a specific procedure without being discharged from the initial hospital– we simplify by only taking sequential admissions into account . Any overlapping admission is treated as having a time between admissions , , of 0 , thus with : ( 4 ) gives the infectious referral rate , per day , between hospitals , where denotes the time span of the dataset . now denotes the probability that any patient will transmit the disease from hospital to within one day . All admissions of all patients combined result in the national hospital network . ( 5 ) ( 6 ) The degree with which hospitals connect with the rest of the hospital network through referrals of patients can be divided into two parts . These consist of the indegree , reflecting the total of introductions a single hospital receives from the rest of the hospital network , and the outdegree which reflects the total amount of colonized patients a single hospital exports to the rest of the hospital network . Because the matrix is asymmetric , and may differ . In order to determine the effect of the difference between inward and outward degree of connectedness , we created a number of datasets with alternative directions . One of these has no direction , the other two have reversed directions . In all three alternatives the university medical centers still have a high degree of connectedness , consistent with the LMR-based network , but a higher outdegree than indegree , contrary to the LMR based network . We first created a dataset without direction , by setting all referral probabilities in the referral matrix equal , but leaving all other parameters the same as the original simulated dataset . We then created a reverse dataset by reversing the time of the original simulated dataset . The new date of admission of a patient , , is simply calculated as , where is the end date of the dataset , in our case day 7300 , and is the discharge date of the patient . This then gives the exact reversion of the original simulated dataset . In order to reverse the direction of the dataset even further , we created another dataset in the same way as the generated dataset with the characteristics of the LMR , in which we set all referral probabilities to university medical centers , in the referral matrix , to zero . This , however , also lowered the overall degree of connectedness of these hospitals . In order to raise the degree we increased the size of the university medical centers 7 fold . The university medical centers now have a higher outdegree than indegree , while their indegree is still higher than the outdegree of the top clinical hospitals . Furthermore , we created a number of small datasets of only five hospitals , in which we varied network properties such as directionality and hospital size ( See Text S1 ) . | The prevalence of hospital acquired infections is widely believed to reflect the quality of health care in individual hospitals , and is therefore often used as a benchmark . Intuitively , the idea is that infections spread more easily in hospitals with a poor quality of health care . This assumes that the rate at which admitted patients introduce new infections is the same for all hospitals . In this article , we show that this assumption is unlikely to be correct . Using national data on patient admissions , we are able to reconstruct the entire hospital network consisting of patients referred between hospitals . This network reveals that university hospitals admit more patients that recently stayed in other hospitals . Consequently , they are more likely to admit patients that still carry pathogens acquired during their previous hospital stay . Therefore , the prevalence of infections does not only reflect the quality of health care but also the connectedness to hospitals from which patients are referred . This phenomenon is missed at the single hospital level; our study is the first to address the connectedness between hospitals in explaining the prevalence of hospital acquired infections . Our findings imply that interventions should focus on hospitals that are central in the network of patient referrals . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"infectious",
"diseases/bacterial",
"infections",
"infectious",
"diseases/nosocomial",
"and",
"healthcare-associated",
"infections",
"infectious",
"diseases/antimicrobials",
"and",
"drug",
"resistance",
"infectious",
"diseases/epidemiology",
"and",
"control",
"of",
"infectious",
"diseases"
] | 2010 | Patient Referral Patterns and the Spread of Hospital-Acquired Infections through National Health Care Networks |
The peptidoglycan of Staphylococcus aureus is characterized by a high degree of crosslinking and almost completely lacks free carboxyl groups , due to amidation of the D-glutamic acid in the stem peptide . Amidation of peptidoglycan has been proposed to play a decisive role in polymerization of cell wall building blocks , correlating with the crosslinking of neighboring peptidoglycan stem peptides . Mutants with a reduced degree of amidation are less viable and show increased susceptibility to methicillin . We identified the enzymes catalyzing the formation of D-glutamine in position 2 of the stem peptide . We provide biochemical evidence that the reaction is catalyzed by a glutamine amidotransferase-like protein and a Mur ligase homologue , encoded by SA1707 and SA1708 , respectively . Both proteins , for which we propose the designation GatD and MurT , are required for amidation and appear to form a physically stable bi-enzyme complex . To investigate the reaction in vitro we purified recombinant GatD and MurT His-tag fusion proteins and their potential substrates , i . e . UDP-MurNAc-pentapeptide , as well as the membrane-bound cell wall precursors lipid I , lipid II and lipid II-Gly5 . In vitro amidation occurred with all bactoprenol-bound intermediates , suggesting that in vivo lipid II and/or lipid II-Gly5 may be substrates for GatD/MurT . Inactivation of the GatD active site abolished lipid II amidation . Both , murT and gatD are organized in an operon and are essential genes of S . aureus . BLAST analysis revealed the presence of homologous transcriptional units in a number of gram-positive pathogens , e . g . Mycobacterium tuberculosis , Streptococcus pneumonia and Clostridium perfringens , all known to have a D-iso-glutamine containing PG . A less negatively charged PG reduces susceptibility towards defensins and may play a general role in innate immune signaling .
In gram-positive bacteria a thick multilayered peptidoglycan ( PG ) constitutes the major component of the cell wall and is essential for survival , maintenance of cell shape and counterbalance of turgor pressure [1] . The heteropolymer consists of alternating disaccharide units composed of N-acetyl-glucosamine ( GlcNAc ) and N-acetyl-muramic acid ( MurNAc ) , that are crosslinked by short peptides to form a rigid network . The biosynthesis of PG is a multistep process which requires numerous enzymatic reactions , occurring in three compartments of a bacterial cell; the cytoplasm ( synthesis of nucleotide-bound precursors ) , the inner face of the membrane ( synthesis of the cell wall building block lipid II and lipid II modifications ) and the outer face of the membrane ( polymerisation of lipid II into the growing PG network ) . Biosynthesis starts in the cytoplasm , where the MurA-F ligases catalyze the formation of the ultimate soluble cell wall precursor UDP-MurNAc-pentapeptide [2]–[5] . In the following membrane associated step , UDP-N-acetyl-muramic acid-pentapeptide is linked to the membrane carrier undecaprenol-phosphate ( C55-P ) by the translocase MraY , resulting in the formation of lipid I ( undecaprenylphosphate-MurNAc-pentapeptide ) . MurG subsequently links UDP-GlcNAc to the N-acetyl-muramic acid moiety of lipid I , yielding lipid II ( undecaprenylphosphate-GlcNAc-MurNAc-pentapeptide ) [3] , [6] , [7] . In Staphylococcus aureus this central cell wall building block is further modified by the addition of a pentaglycine interpeptide bridge , catalyzed by the FemXAB peptidyltransferases [8]–[12] . The modified lipid II is then translocated across the cytoplasmic membrane engaging FtsW flippase activity [13] and assembled into the growing PG network , through the activity of penicillin binding proteins ( PBPs ) by transglycosylation and transpeptidation reactions [14]–[16] . Following synthesis and assembly the S . aureus PG undergoes further modifications including O-acetylation of N-acetyl-muramic acid [17]–[19] and the addition of structures that are covalently linked , such as wall teichoic acids [20] , proteins and capsules [21] , [22] . Apart from this , the PG of staphylococci almost completely lacks free carboxyl groups , since the α-carboxyl group of D-glutamic acid at position 2 of the stem peptide is amidated , resulting in the formation of D-iso-glutamine [23] . Using a cell-free system with crude membrane preparations Siewert and Strominger ( 1968 ) suggested that the lipid bound precursors could serve as acceptors of ammonia in an ATP-dependent reaction [24] . Until now the primary role of D-Glu amidation of the stem peptide has remained elusive and the enzyme catalyzing the amidation reaction has not been identified so far . Extensive genetic analysis and characterization of mutant cell walls revealed several loci in the genome of S . aureus affecting the degree of muropeptide amidation [25] , [26] . A femC ( factor essential for methicillin resistance ) mutant was described exhibiting 48% decreased muropeptide amidation [27] , accompanied by a reduction in methicillin resistance . The femC phenotype resulted from the disruption of the glutamine synthetase repressor glnR causing a polar effect on glutamine synthetase ( GlnA ) expression , which in turn led to a drastic reduction of the intracellular glutamine pool [27] . Together with the observation that the external addition of glutamine to the medium restored the femC defect [28] , this could suggest that glutamine may be a nitrogen donor . On the functional level D-Glu amidation appears to be relevant for efficient transpeptidation of neighboring stem peptides . A considerable reduction of PG crosslinking was observed in the femC mutant accompanied by a high percentage of free D-Ala-D-Ala termini [11] , [29] , suggesting that non-amidated cell wall precursors are imperfect substrates for one or more transpeptidases . In accordance , it has been shown early by Nakel et al . that crosslinking of two adjacent stem peptides requires at least one of the stem peptides involved to be amidated [30] . Characteristically , the PG of S . aureus is extensively crosslinked , with up to 95% of the stem peptides interconnected [31] . The coordinated crosslinking therefore plays a decisive role for S . aureus survival; its perturbation appears to impair growth and to cause aberrant septum formation and severe cell deformation . Interestingly , FemC is essential for the full expression of methicillin resistance in methicillin-resistant S . aureus ( MRSA ) , as also observed with femXAB mutants which are defective in pentaglycine bridge formation [32] . In this study , we identified the enzymes catalyzing stem peptide amidation . We found the glutamine amidotransferase-like protein SA1707 ( designated GatD; UniProtKB: Q7A4R4 ) in concert with the Mur ligase homologue SA1708 ( designated MurT; UniProtKB: Q7A4R3 ) to catalyze the amidation of the α-carboxyl group of D-glutamic acid of cell wall precursor stem peptides . Characterization of these enzymes and in vitro reconstitution of their amidation reaction will enable detailed analysis of their functional role on transpeptidation and translocation of cell wall precursors as well as provide new opportunities to identify selective antibiotic inhibitors of these essential proteins .
S . aureus COL and S . aureus RN4220 were maintained on blood or Luria Bertani ( LB ) broth ( Oxoid ) . S . aureus strains carrying antisense plasmids ( kindly provided by Merck ) were maintained on LB-agar plates supplemented with 34 µg/ml chloramphenicol [33] , [34] . E . coli BL21 was used for overexpression of recombinant His6-tag fusion proteins and maintained on LB-agar plates containing 50 µg/ml ampicillin . MRSA COL ( MB 5393 ) was transformed with antisense interference plasmids correspond to SA1707 ( SAV1891 ) , SA1708 ( SAV1892 ) or vector control as previously described [35] . Assay plates were prepared by seeding 107 cells/ml of each culture into 48°C cooled LB Miller agar containing 34 µg/ml chloramphenicol with or without 50 mM xylose . Agar plates were allowed to set and then spotted with 10 µl of each drug as previously described [36] and incubated at 37°C with humidity for 18 hours . Plectasin was tested at increasing amounts ranging from 0 . 5 to 10 µg and spotted in 10 µl aliquots each . S . aureus N315 murT ( SA1708 ) , gatD ( SA1707 ) and glnA were amplified using forward and reverse primers as listed in Table 1 and cloned into a pET21b vector ( Novagen ) using NdeI and XhoI restriction sites to generate C-terminal His6-fusion proteins . E . coli BL21 ( DE3 ) ( Promega ) cells transformed with the appropriate recombinant plasmid were grown in LB-medium ( 50 µg/ml ampicillin ) at 37°C . At an OD600 of 0 . 6 , IPTG was added at a concentration of 0 . 75 mM to induce expression of the recombinant proteins . After 4 h , cells were harvested and resuspended in lysis buffer ( 50 mM Tris/HCl , pH 7 . 5 , 300 mM NaCl , 10 mM imidazole ) . Aliquots of 200 mg/ml lysozyme , 100 mg/ml DNase and 10 mg/ml RNase were added; cells were incubated for 30 min on ice and sonicated . Cell debris was spun down . The supernatant was applied to Ni-NTA- ( Qiagen ) or Talon- ( Clontech ) agarose slurry . This mixture was gently stirred at 4°C for 2 h and then loaded onto a column support . After washing with lysis buffer , weakly bound material was removed with 50 mM Tris/HCl , pH 7 . 5 , 300 mM NaCl and 20 mM imidazole . His-tagged recombinant proteins eluted with buffer containing 50 mM Tris/HCl , pH 7 . 5 , 300 mM NaCl and 100–200 mM imidazole . Three fractions were collected each and were stored in 30% glycerol at −20°C . Purity was controlled by SDS-Page . The gatD/murT operon was amplified using primers murT-for and gatD-rev ( Table 1 ) and cloned into pET21b vector ( Novagen ) generating a C-terminal gatD-His tag fusion . The corresponding plasmid was introduced into E . coli BL21 ( Promega ) and overexpression was carried out as described above . Co-elution from Ni-NTA column ( Qiagen ) of the GatD-His6/MurT complex was analyzed by SDS page . Site directed mutagenesis of GatD catalytical triad/active site was carried out using the QuikChange Lightning Mutagenesis Kit ( Stratagene ) following the instructions of the manufacturer using plasmid pET21-gatD as the template . Mutagenesis primers ( GatD_mut1; GatD_mut2 ) are listed in Table 1 , resulting in exchange of Cys ( TGT ) >Gly ( GGA; 348 ) . The GatD_mut ( C94G ) protein has been purified as described above . [14C]-N-acetyl muramic acid pentapeptide ( UDP-MurNAc-pp ) was synthesized on the basis of the protocol described by Wong et al . elaborated for E . coli [37] with modifications [38] . In short , 100 nmol UDP-GlcNAc were incubated in the presence of 2 µg MurA-F and DdlA protein each in 50 mM Tris-Bis-propane , pH 8 , 25 mM ( NH4 ) 2SO4 , 5 mM MgCl2 , 5 mM KCl , 0 . 5 mM DTT , 2 mM ATP , 2 mM PEP , 2 mM NADPH , 1 mM of each amino acid ( L-Lys , D-Glu , L-Ala , D-Ala , respectively ) and 10% DMSO in a total volume of 125 µl for 90 min at 30°C . If not mentioned elsewhere the radiolabel was introduced using [14C]-L-lysine . Purification was performed as described [39] . Lipid I-synthesis was carried out in a total volume of 60 µl containing 2 . 5 nmol C55-P , 25 nmol of UDP-MurNAc-pp in 100 mM Tris-HCl , 30 mM MgCl2 , pH 7 . 5 , and 10 mM N-lauroyl sarcosine . The reaction was initiated by the addition of 7 . 5 µg of the enzyme and incubated for 90 min at 30°C . Synthesized lipid I was extracted from the reaction mixture with n-butanol/pyridine acetate , pH 4 . 2 ( 1∶1; v/v ) and purification and quantification was carried out as described for lipid II [12] . [14C]-labeled lipid I was synthesized in the presence of 25 nmol [14C]-UDP-MurNAc-pentapeptide . Synthesis and purification of lipid II was performed using membranes of Micrococcus luteus as described [40]–[42] . In short , membrane preparations ( 200 µg protein ) were incubated in the presence of purified substrates , 5 nmol undecaprenylphosphate ( C55-P ) , 50 nmol UDP-MurNAc-pp and 50 nmol [14C]-UDP-GlcNAc in 60 mM Tris-HCl , 5 mM MgCl2 , pH 7 . 5 , and 0 . 5% ( w/v ) Triton X-100 in a total volume of 50 µl for 1 h at 30°C . Bactoprenol containing products were extracted with the same volume of butanol/pyridine acetate ( 2∶1; vol∶vol; pH 4 . 2 ) and analyzed by TLC using phosphomolybdic acid ( PMA ) staining . For synthesis of mg-quantities of lipid II the analytical assay was scaled up and purification was performed as described [12] . Lipid II-Gly5 was synthesized using purified lipid II , purified recombinant FemXAB peptidyltransferases , tRNA and Glycyl-tRNA-synthetase as described previously [12] . Purification was performed as described for lipid II . The assays for synthesis of amidated lipid intermediates were performed in a total volume of 30 µl containing 2 µg of purified MurT-His6 and GatD-His6 . If not indicated elsewhere , 2 nmol of purified lipid intermediates , lipid I and lipid II , respectively were incubated in 160 mM Tris-HCl , 0 . 7% Triton X-100 , 5 mM KCl , 40 mM MgCl2 , pH 7 . 5 , 6 mM ATP and 7 mM glutamine for 2 h at 30C° . Synthesis products were extracted from the reaction mixture with the same volume of n-butanol/pyridine acetate , pH 4 . 2 , and analyzed by TLC solvent B ( butanol , acetic acid , water , pyridine , 15∶3∶12∶10 ) . Radiolabeled spots or lanes were visualized using a storage phosphor screen in a Storm imaging system ( GE Healthcare ) . Non-radiolabeled lipid intermediates were analyzed using PMA staining . Isolation of larger quantities of non-radioactive-labeled amidated lipid II intermediates was achieved with an upscale of the synthesis assay and subsequent purification via preparative TLC . To this end , lipid spots were visualized using iodine vapor and material was scratched of the silica plates . Lipids were extracted by incubation in 100 µl of chloroform/methanol ( 1∶1; v/v ) for 60 min . Electrospray MS spectra were recorded on a micrOTOF-Q quadrupole-TOF instrument ( Bruker ) working in negative mode . Samples were infused at 0 . 2–3 ml h-1 , either directly ( in methanol–chloroform , 1∶1 ) or diluted 1∶1 in methanol . The spectra were externally calibrated with sodium formiate in methanol .
S . aureus membrane preparations possessing the enzymatic activity of MraY and MurG synthesize cell wall precursors lipid I and lipid II , respectively [43] . Furthermore Siewert and Strominger showed that after addition of ATP , NH4Cl or glutamine , amidated lipid I or lipid II can be detected in such membranes [24] . We used purified lipid II together with S . aureus membranes , glutamine and ATP and observed an additional lipid II band , distinguished by an elevated Rf-value ( Figure 1 , lane 2 ) . When glutamine and ATP were omitted from the reaction mixture , predominantly unmodified lipid II was detected ( Figure 1 , lane 3 ) ; marginal conversion to the newly formed lipid II ( lane 3 ) might result from traces of residual ATP and glutamine in the membrane preparation . Therefore , lipid II appears to be a direct substrate for this modification , as initially proposed [24] . The newly formed modified lipid II was analyzed by ESI-TOF-MS running in negative mode ( Figure 2 ) . A mass decrease of 1 ( 0 . 5 for the doubly charged ion ) is consistent with amidation of the α-carboxyl-OH group ( Figure 2 ) . The altered migration behavior of amidated lipid II in the TLC-system provided a convenient and robust assay for further analysis of the amidation reaction . Recently , a large-scale antisense interference-based phenotypic screen was performed to identify genes required for the broad β-lactam resistance characteristic of methicillin-resistant Staphylococcus aureus ( MRSA ) [36] . Two yet uncharacterized open reading frames identified in this analysis correspond to SA1707 and SA1708 . Upon antisense-mediated depletion in expression of SA1707 and SA1708 , MRSA strain COL displayed prominently restored susceptibility to diverse carbapenem and cephalosporin β-lactam antibiotics ( Figure S1 ) . Moreover depletion resulted in increased susceptibility to plectasin ( Figure S2 ) , a defensin known to inhibit cell wall biosynthesis through specific binding of lipid II [44] . SA1707 encodes a putative glutamine amidotransferase with homology to cobyric acid synthases , whereas the co-transcribed SA1708 gene ( Figure S3 A ) encodes an uncharacterized protein with homology to the Mur-ligases MurE and F , involved in pentapeptide side chain assembly during peptidoglycan synthesis . Progressive reduction of SA1707 and SA1708 expression through increasing xylose concentrations impaired the growth rate , strongly suggesting that both genes are essential for S . aureus viability ( Figure S4 ) . Nevertheless morphology of partially GatD/MurT depleted cells was unaltered , as revealed by electron microscopy ( data not shown ) . BLAST analysis revealed the presence of an equivalent gene arrangement in a number of gram-positive bacteria , such as M . tuberculosis , S . pneumonia and C . perfringens ( Figure S3 B ) . Interestingly , only bacteria which are reported to contain an amidated PG [23] , encode homologues of SA1707 and SA1708 , suggesting their potential functional role in PG amidation ( Figure S3 ) . Accordingly , based on sequence similarities to glutamine amidotransferases ( GATases ) and Mur ligases , as well as in vitro biochemical evidence ( see below ) , we propose to designate SA1707 and SA1708 as GatD and MurT , respectively . To investigate GatD and MurT function both proteins were purified as His-tag fusion proteins ( purity >95% ) and an individual in vitro assay was set up based on the information obtained using membrane preparations ( Figure 3A ) . As revealed by TLC , neither MurT , nor GatD alone were sufficient to catalyze the amidation of lipid II when added separately to the reaction mixture ( lane 3 and 4 ) , as no change in migration behavior was observed compared to the negative control ( lane 1 ) . However co-addition of GatD and MurT to the assay ( 2 µg each ) resulted in complete conversion of lipid II to an amidated lipid II species ( lane 2 ) . Omission of glutamine from the reaction mixture resulted in no formation of amidated lipid II ( lane 5 ) , further providing evidence for glutamine to be a direct nitrogen donor for cell wall precursor amidation . Monitoring the reaction over time showed a maximum glutamine-dependent conversion to the reaction product after incubation for 2 h ( Figure 3B ) with a pH optimum of 7 . 5–7 . 8 . At pH 5 . 5 GatD/MurT were completely inactive ( data not shown ) . To investigate the co-requirement of GatD and MurT for lipid II amidation , MurT was individually substituted by S . aureus Mur-ligases MurC-F in the synthesis assay , as exemplarily shown with MurE ( Figure 4 A ) . However , in spite of sharing sequence identity of up to 23% with S . aureus Mur ligases , MurT functional activity could not be replaced in the in vitro lipid II amidation assay using purified MurC , D , E and F proteins . Interestingly , MurT could substitute for MurE activity in vitro , resulting in the formation of UDP-MurNAc- ( Ala-Glu-Lys ) -tripeptide ( data not shown ) . Despite these results , however , it appears unlikely that S . aureus MurT is able to substitute for MurE in vivo , since murT or murE are each essential ( unlike for example the non-essential murA , murZ paralogs ) [45]–[48] . Further a murE-transposon mutant with reduced specific activity was shown to accumulate UDP-MurNAc-dipeptide in the cytoplasm [49] , [50] . Incubation of increasing concentrations of MurT ( 0–2 µg ) in the presence of 2 µg GatD ( Figure 4 B; squares ) further substantiated the interdependency of both proteins . Increasing MurT concentrations led to enhanced amidation of lipid II with a maximum activity observed at 1 . 5 µg of MurT , which corresponds to a molar ratio of MurT∶GatD of 1∶1 ( molecular masses of 49 . 2 kD for MurT and 27 . 4 kD for GatD ) , suggesting the formation of a heteromeric complex by the two proteins . We then co-expressed both genes with a His-tag attached only to the C-terminus of GatD . The Ni-NTA column eluate contained both enzymes in similar amounts ( Figure S5 ) , strongly suggesting that in vivo both enzymes form a physically stable bi-enzyme complex . Analysis of sequence similarity identified GatD as a member of the superfamily of glutamine amidotransferases ( GATases ) . These enzymes catalyze the transfer of an amide nitrogen from glutamine to its substrate to form a new carbon-nitrogen group [51] . Until now 16 glutamine amidotransferases have been identified , which are grouped into two classes: class-I ( also referred to as trpG-type ) and class-II ( also known as purF-type ) [52] . GATases possess two functional domains; a glutaminase- and a synthase- domain , which may either be expressed as a single protein or separate subunits which form a heterodimeric GATase complex [51] , [52] . The 243 amino acid GatD protein specifically shares sequence similarities with glutaminase domains of class I-type GATases , that hydrolyse glutamine to generate glutamate and ammonia ( NH3 ) [53] . S . aureus GatD shares the conserved cysteine and histidine residues of class I-type GATases . In the bacterial enzymes described here the third catalytical triad residue glutamine is missing , however a glycine residue appears highly conserved ( Figure 5 B ) . Within the active site the cysteine is essential for glutaminase activity [52]–[54] , since its nucleophilic sulfhydryl side chain initiates the amide transfer through the formation of a thioester with the substrate glutamine [55] . To explore the functional role of catalytical triad , site-directed mutagenesis of S . aureus GatD was performed by replacing the canonical cysteine of the proposed active site ( position 94 ) with glycine ( Figure 5 B ) . Unlike wild type GatD ( Figure 5 A , lane 2 ) , amidation of lipid II was not observed using the GatD_C94G protein . The inability of the GatD mutant to use glutamine provides further evidence for the function of GatD as a glutaminase and confirms the active site of the enzyme . Consistent with these findings , the catalytical triad active site and structural features of S . aureus GatD appear broadly conserved amongst gram-positive bacteria , including M . tuberculosis , C . perfringens and S . pyogenes ( Figure 5 B ) . As GatD and MurT enzymes catalyze the amidation of lipid II ( Figure 3 ) , we further included purified lipid I and lipid II-Gly5 in the in vitro assay , in attempt to narrow down the primary acceptor substrate and stage of amidation . Amidation analysis revealed that both lipid intermediates , lipid I and lipid II-Gly5 , also serve as a substrate ( Figure 6 ) and complete conversion to the amidated lipid variant was found when incubated in the presence of ATP and glutamine . Conversely , addition of purified UDP-MurNAc-pentapeptide at 10-fold molar excess with respect to lipid II present in the synthesis assay , had only a minor impact on the formation of amidated lipid II; no influence on lipid II amidation was observed upon UDP-MurNAc-pentapeptide addition reduced at 5-fold molar excess ( Figure 6 ) . These data support the conclusion that amidation exclusively occurs at the stage of bactoprenol-bound cell wall precursors . To investigate the possibility of a concerted activity with the MurC-F enzymes of S . aureus during stem peptide formation , we incubated MurT and GatD in the presence of purified MurA-F enzymes , UDP-GlcNAc , ATP and glutamine . Following inactivation of these enzymes , the reaction products were incubated with C55-P and MraY for another 60 minutes and reaction products were analyzed by TLC . As shown in Figure 6 B no change in the migration behavior was observed when MurA-F enzymes were incubated in the presence of GatD/MurT ( lane 4 ) , compared to the control where GatD/MurT were omitted and lipid I was formed . Moreover , either in the presence ( lane 4 ) or absence of glutamine ( lane 5 ) , only the formation of unmodified lipid I was detected , suggesting that amidation does not occur during MurC-F catalyzed stem peptide formation and that the soluble cell wall precursor UDP-MurNAc-pentapeptide does not serve as a substrate for the amidation reaction . Moreover , these results are in good agreement with the fact that only non-amidated UDP-MurNAc-pentapeptide was isolated from the cytoplasm of S . aureus and from other staphylococcal strains used in this study to purify UDP-MurNAc-pentapeptide , as analyzed by mass spectrometry ( data not shown ) . To investigate the donor substrates of GatD/MurT , the in vitro assay was supplemented with various potential nitrogen donors . As summarized in Table 2 , GatD/MurT exclusively utilize glutamine as the nitrogen donor at neutral pH and neither ammonia nor NH4Cl were found to be substrates . Conversely , in combination with purified glutamine synthase GlnA , amidation was observed in the presence of glutamate and NH4Cl , resulting from the GlnA catalyzed conversion of glutamate and NH4Cl to glutamine , the latter of which then serves as a substrate for GatD/MurT . In contrast , at pH 8 . 5 where the concentration of unprotonated ammonia is higher , NH3-dependent activity of GatD/MurT was also observed in vitro , a finding that has been reported earlier for other GATases [56] . Interestingly , with the GatD_C94G mutant enzyme NH3-dependent activity of MurT was unaffected , while a complete loss of glutamine-dependent activity was observed , suggesting different binding sites for glutamine and NH3 on the two subunits . Again , neither MurT nor GatD alone were found to catalyze the amidation of lipid II independently , irrespective of the pH or the nature of the nitrogen donor , further demonstrating that the concerted action of GatD/MurT is a prerequisite for amidotransferase activity .
Cell wall biosynthesis is a vital and highly dynamic process for almost all bacteria requiring continuous biosynthesis and maintenance involving species-specific modifications . Among these modifications the amidation of the peptidoglycan ( PG ) constitutes a relatively minor biochemical variation , but is of central importance for S . aureus viability . Until now the primary role of the amidation of D-glutamic acid in position 2 of the stem peptide has remained elusive and an enzyme catalyzing the reaction has not been identified so far . In this study we identified GatD ( SA1707 ) and MurT ( SA1708 ) as the enzymes catalyzing the amidation of the S . aureus peptidoglycan building block . In vitro analysis using purified proteins and substrates demonstrated that amidation is catalyzed by a glutamine amidotransferase bi-enzyme complex encoded by two so far uncharacterized open reading frames SA1707 ( GatD ) and SA1708 ( MurT ) . Glutamine amidotransferases ( GATases ) in general are involved in a variety of cellular processes like synthesis of amino acids , nucleotides , amino sugars , and antibiotics [51] . Characteristically GATases are composed of two different catalytic domains , each contributing to the catalysis of a single biochemical reaction; hydrolysis of glutamine ( glutaminase domain ) and the transfer of reduced nitrogen to its specific acceptor substrate ( synthase domain ) . Both reactions are tightly coupled and require the close interaction of both folding domains , that can either derive from a single polypeptide , from two distinct polypeptides or more infrequently derive from different enzymes which then form a heterodimeric GATase [52 and references therein] , as demonstrated here for GatD ( glutaminase ) and MurT ( ATP-dependent synthetase ) . Most GATases have been reported to use both glutamine and NH3 as a nitrogen donor and to contain two corresponding nitrogen substrate binding sites , a glutamine- and an ammonia-dependent binding site [52] . This is in line with our finding that at pH 8 . 5 , where the concentration of unprotonated ammonia is higher , free ammonia also served as a substrate for the GatD/MurT catalyzed reaction in vitro . In contrast to the glutamine-dependent activity , which was abolished by a glycine substitution in place of the catalytical cysteine residue , the ammonia-dependent amidation of lipid II was unaffected when MurT was incubated in the presence of the GatD_C94G enzyme . Although the enzymatic function of the glutaminase domain is dispensable when NH3 is used as nitrogen donor at pH 8 . 5 , the interaction of MurT with its cognate glutaminase domain appears to be crucial for the overall GATase activity . Several conformational changes have been reported to occur upon domain assembly to form the active enzyme for a number of different GATases from several organisms [57]–[59] . As shown for glutamine synthase GlmS , specific residues belonging to the synthase domain participate in the glutaminase site and thereby contribute to the coupling of the active sites of both domains [60] . As further observed by structural analysis , most GATases shuttle ammonia through a solvent-inaccessible channel from the glutaminase-active site to the synthase-active site [57] , [61] , to prevent NH3 release and the formation of non-reactive ammonium ions as well as toxic side effects within the cell . These ammonia channels are predominantly formed by the synthase domain and its formation has been shown to also require the presence of the acceptor substrate in most GATases [57] , [61] . Moreover , both glutaminase- and synthase-activities are only optimal in the presence of nitrogen donor and acceptor substrate , ensuring a functional coordination [52] . Such an interdomain signaling mechanism might ensure regular substrate binding and consumption . Likewise synchronization might promote the efficiency of the reaction that with regard to the essentiality of the acceptor substrate lipid II , needs to be highly coordinated to allow for a proper processing of cell wall biosynthesis . The amidation reaction appears highly specific for glutamine and no other nitrogen donor was accepted by GatD/MurT except for ammonia at elevated pH . Considering the neutral pH within the cytoplasm and the fact that in a FemC mutant , incapable of synthesizing glutamine , the amidation of PG is dramatically reduced , it is most likely that in vivo glutamine serves as the nitrogen donor for GatD/MurT . In contrast to the glutaminase domain of GATases , which are highly homologous throughout the entire GAT-family , the synthase domains differ largely as to the different acceptor substrates used and the different underlying biochemical reactions catalyzed [52] . MurT shares sequence similarity with the Mur-ligase MurE . Mur-ligases ( C-F ) catalyze the sequential assembly of the pentapeptide side chain of the soluble cell wall precursor UDP-MurNAc-pentapeptide . Mechanistically , these reactions proceed via the formation of an acyl phosphate intermediate by phosphorylation of the respective C-terminal carboxylate of the UDP-activated sugar at the expense of one molecule of ATP [62] . As shown for various GATases the amidation reaction itself is independent of ATP hydrolysis , while for a subgroup of enzymes the ATP-dependent activation of the acceptor substrate has been described [52] . The GatD/MurT catalyzed reaction is ATP- and Mg2+-dependent , strongly suggesting the formation of an activated acceptor intermediate prior to amidation , in which the phosphate group is then displaced by the incoming nitrogen group . This is further supported by the observation that MurT was able to substitute MurE in an in vitro UDP-MurNAc-pentapeptide synthesis assay , assuming the capability to bind L-lysine , UDP-MurNAc-dipeptide and to activate the D-Glu carboxylate by phosphorylation . As deduced from a sequence alignment with MurE ligase , MurT exhibits the ATP consensus binding site GTNGKT and a short sequence ( DNAADD ) with similarity to the L-Lys binding site ( DNPAND ) present in MurE [63] , [64] . Nevertheless , since MurE and MurT are both essential , MurT appears unable to normally substitute for MurE function in vivo . Despite the ability of MurT to functionally substitute for MurE-dependent stem peptide synthesis in vitro , we propose MurT-catalyzed amidation likely occurs at the stage of bactoprenol-bound cell wall precursors , prior to their export to the outside and subsequent polymerization into the growing PG network ( Figure 7 ) . This is based on the observation that lipid I , lipid II , and lipid II-Gly5 are all amidated in a GatD/MurT dependent manner , whereas UDP-MurNAc-pentapeptide , did not interfere with the GatD/MurT catalyzed amidation reaction . However , as all lipid intermediates were found to serve as a substrate for GatD/MurT in vitro to the same extent under the conditions chosen , additional experiments are required to further elucidate the sequence of the reaction in vivo . Recently MraY and MurG have been reported to form a complex [65] , [66] , suggesting a sterically unfavorable situation for the amidation of lipid I . Further , in the protist Cyanophora paradoxa modification of the D-Glu of the stem peptide with putrescine has also been reported to occur at a membrane-bound stage of peptidoglycan biosynthesis and appeared to be more efficient with lipid II as a substrate [67] . Considering that GatD/MurT are inactive at pH 5 . 5 we therefore presume that amidation likely occurs after lipid II or lipid II-Gly5 is formed ( Figure 7 ) . Functionally , amidation could facilitate the translocation of the cell wall building block lipid II across the cytoplasmic membrane as a consequence of the reduction of polarity . Amidation actually may provide a signal for lipid II translocation , which would then suggest that lipid II-Gly5 serves in vivo as the acceptor , as depicted in the proposed model ( Figure 7 ) . Crosslinking of two adjacent stem peptides via the characteristic pentaglycin-interpeptide bridge requires at least one of the stem peptides involved to be amidated [30] , suggesting that non-amidated lipid II is an inefficient substrate for one or more transpeptidases in S . aureus . This critical role for PG precursor amidation and reduced resistance of MRSA femC mutants to methicillin is also consistent with the broad β-lactam hypersusceptibility phenotypes of GatD and MurT antisense depletion strains we observed ( Figure S1 ) . Considering that PG precursor translocation , transglycosylation , and transpeptidation reactions are tightly interlinked , GatD/MurT dependent amidation may function to coordinate these biochemical events within this macromolecular heteromeric complex . For example , synthetically generated amidated muropeptides have recently been shown to be preferred substrates for the Ser/Thr kinase PknB in M . tuberculosis compared to their non-amidated counterparts [68] . Interestingly , the extracytoplasmic part of the membrane anchored PknB protein comprises repeating units of PASTA domains ( penicillin binding protein and Ser/Thr kinase associated ) , predicted to bind to the D-Ala-D-Ala terminus of PG precursors [69] , thus emphasizing the potential relationship between amidation and transpeptidation . Amidation may also constitute a checkpoint for PknB dependent PG turnover , cell growth and cell division . Interestingly , like GatD and MurT , PknB antisense depletion strains were also identified to display dramatically restored β-lactam susceptibility phenotypes among MRSA isolates , emphasizing their common participation in PG biosynthesis and cell wall biogenesis . Since amidation of D-Glu also results in a less negatively charged PG , reduction of susceptibility towards innate defense mechanisms , provided by cationic molecules such as defensins and lysozym is very likely . In line with this , we observed increased efficacy of plectasin against GatD/MurT depleted cells . Plectasin has been shown to specifically bind to lipid II and its N-terminal amino group is supposed to contribute to binding through interaction with the carboxyl group of the D-Glu residue [44] . Considering the essentiality of these proteins in S . aureus ( Figure S4 ) [46]–[48] , M . tuberculosis [70] and S . pneumoniae [71] , their broad conservation across gram-positive bacterial pathogens , and the development of robust in vitro assays for PG amidation described here , identifying inhibitors to these targets offers a new approach to developing both monotherapeutics as well as combination agents to pair with existing β-lactam antibiotics , thereby restoring their activity against MRSA . | The bacterial peptidoglycan is a hetero-polymer , consisting of sugars and amino acids , that forms a stress-bearing sacculus around bacterial cells and provides cell shape . The cell envelope and its components represent a central interface for interactions with the environment and are therefore subject to species-specific modifications . The peptidoglycan of many Gram positive pathogens such as Staphylococcus aureus is almost fully amidated which appears to reduce the susceptibility towards innate host defenses . Here , we describe the so far elusive enzymes that catalyze the amidation of the peptidoglycan precursors and provide biochemical evidence for acceptor and nitrogen donor substrates . We show that two enzymes are necessary to catalyze the amidation and that both enzymes form a stable heterodimer complex . Besides substantial progress in understanding of peptidoglycan biosynthesis our results provide the molecular basis for screening for mechanistically novel antibiotics . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [
"biotechnology",
"bacteriology",
"medicine",
"biochemistry",
"infectious",
"diseases",
"bacterial",
"diseases",
"biology",
"microbiology",
"bacterial",
"pathogens"
] | 2012 | Identification and in vitro Analysis of the GatD/MurT Enzyme-Complex Catalyzing Lipid II Amidation in Staphylococcus aureus |
Dengue is a growing global concern with 390 million people infected each year . Dengue virus ( DENV ) is transmitted by mosquitoes , thus host cells in the skin are the first point of contact with the virus . Human skin contains several populations of antigen-presenting cells which could drive the immune response to DENV in vivo: epidermal Langerhans cells ( LCs ) , three populations of dermal dendritic cells ( DCs ) , and macrophages . Using samples of normal human skin we detected productive infection of CD14+ and CD1c+ DCs , LCs and dermal macrophages , which was independent of DC-SIGN expression . LCs produced the highest viral titers and were less sensitive to IFN-β . Nanostring gene expression data showed significant up-regulation of IFN-β , STAT-1 and CCL5 upon viral exposure in susceptible DC populations . In mice infected intra-dermally with DENV we detected parallel populations of infected DCs originating from the dermis and migrating to the skin-draining lymph nodes . Therefore dermal DCs may simultaneously facilitate systemic spread of DENV and initiate the adaptive anti-viral immune response .
Aedes mosquitoes are the primary vectors for the transmission of dengue virus ( DENV ) . While probing for blood microvessels from which to feed , the mosquito releases virus-containing saliva into the dermal layer of the skin . Studies using mosquitoes infected with the closely-related West Nile virus showed that more than 99% of the viral particles could be recovered from around the feeding site on mice , indicating that most of the virus is not injected directly into the blood but rather pools in the local tissue [1] . Precisely how such viruses , including West Nile and DENV , then spread to cause systemic infection is currently unknown . Human skin is composed of an epidermal and a dermal layer , separated by the basement membrane . The epidermis contains keratinocytes and Langerhans Cells ( LCs ) , a specialized type of dendritic cell ( DC ) that constantly probes for antigen in the most exposed , superficial layer of the skin [2] . Upon detection of pathogens during an infection LCs migrate to draining lymph nodes ( LNs ) where they contribute to the initiation of T cell responses . Although early studies suggested that LCs were the principal migratory DC initiating T cell responses , more recent findings have demonstrated a key role for LCs in Treg activation and skin homeostasis [3]–[5] . Mice in which LCs have been depleted still generate protective skin-specific T cell responses [6] . The underlying dermis , in contrast , contains fibroblasts as wells as high numbers of immune cells including macrophages , T cells and three subsets of dendritic cells ( DCs ) [7]–[9] . Additionally , the dermis harbors a dense network of blood and lymphatic vessels [7] , through which immune cells and mediators can both enter and exit the tissue . The three dermal DC subsets are distinguished by positive expression of CD1c ( a MHC I-related molecule that presents lipids to T cells ) , CD14 ( co-receptor for bacterial lipopolysaccharide ) or CD141 ( thrombomodulin ) . CD1c+ DCs are the most abundant amongst the three subsets , and following activation in the skin their functional role is to migrate to draining LNs for the initiation of systemic T cell responses [10] . CD14+ DCs are less abundant than CD1c+ DCs , and were recently shown to be monocyte-derived cells transcriptionally related to macrophages [11] [12] . Skin CD14+ DCs have the capacity to activate CD4+ T cells and drive their differentiation into T follicular helper cells ( Tfh ) that support the efficient initiation of antibody responses [10] , [13] . CD14+ DCs also have the capacity to induce tolerance by promoting the generation of Tregs in the presence of Vitamin D3 [5] , [14] . The third and least abundant subset of skin DCs are the CD141+ DCs , which also migrate to LNs but specialize in cross-presenting antigen to CD8+ T cells [15] . Recently , murine homologs of human tissue DC subsets were identified [9] , [15] , which raises the possibility of drawing new parallels from findings using murine viral infection models . In this study we interrogated the host target cells of DENV at the physiological entry site of infection in human skin , to understand their functional relevance in the development of dengue-specific infection and immunity . Our findings demonstrate heterogeneity in susceptibility to dengue virus infection within skin APC subsets in both humans and mice . These results enhance our understanding of the early consequences of dengue virus infection .
To identify the cell types within human skin that are susceptible to DENV infection we prepared single cell suspensions from healthy skin obtained from mastectomy or abdominoplasty surgery , and exposed the cells to DENV-2 strain D2Y98P [16] at an MOI of 2 . After 48 h , flow cytometry was used to characterize the infected cell types by measuring the percentage of cells positive for DENV E protein , which forms part of the viral envelope ( Fig . 1 ) . The vast majority ( approximately 90% ) of cells positive for E protein expressed CD45 and HLA-DR ( Fig . 1A and B ) . This finding excluded significant infection of CD45− keratinocytes , fibroblasts and endothelial cells , which had been reported previously to be possible targets of DENV infection [17] , [18] . CD45+ HLA-DR+ cells include all antigen-presenting cells ( APCs ) in the skin . To further refine our analysis we employed a previously described gating strategy to distinguish between CD14+ DCs , CD1c+ DCs , CD141+ DCs and LCs ( [15] and S1A Figure ) . Three skin DC subsets were susceptible to DENV-2 infection ex vivo: LCs and CD14+ DCs were infected most efficiently while CD1c+ cells showed a lower infection rate ( Fig . 1C ) . To prove that E protein-positive cells were truly infected and had not only taken up virus particles , cells were also treated with UV-inactivated virus as a control ( Fig . 1C ) . The infection profile for skin DC subsets was reproducible and independent of the skin donor ( Fig . 1D ) . Interestingly , we did not detect infection of CD141+ DCs . To test whether there were serotype-specific differences in infection we also exposed single cell suspensions from skin to DENV-1 , -3 and -4 . These experiments showed that LCs and CD14+ DCs were consistently infected at higher rates than CD1c+ and CD141+ DCs , showing a similar DENV infection profile in human skin cells to DENV-2 ( Fig . 1E ) . Of note , infection with DENV-1 , -3 and -4 was less efficient than infection with DENV-2 D2Y98P ( Fig . 1A and B ) , which was expected due to the enhanced viral RNA synthesis capacity of the latter [19] . A less virulent DENV-2 strain ( TSV01 ) showed lower infection rates than D2Y98P , but a similar target cell infection profile ( Fig . S1C ) . In addition to DCs we identified dermal macrophages by flow cytometry based on their auto-fluorescence in the FITC channel [20] ( Fig . S1A and B ) . CD14+ DCs and dermal macrophages were infected at similar rates 24 h after infection ( Figure S1B ) , identifying both DCs and macrophages in the skin as potential DENV targets . As Aedes mosquitoes deposit virus-containing saliva in the dermal layer , bypassing the epidermis when probing with their proboscis to search for blood vessels , we explored infection susceptibility upon intradermal delivery of DENV . LCs were not infected when the virus was injected intra-dermally , in contrast to infection of skin single cell suspension ( Figure S1C ) . Whereas flow cytometry only detects how much viral protein is expressed in infected cells , production of infectious virus particles by infected cells can be assessed by using a plaque-forming assay or by measuring viral RNA by PCR in the cell culture supernatant . To further characterize the infection kinetics of various skin DC subsets we infected single cell suspensions of human skin with DENV-2 strain D2Y98P at an MOI of 2 , analyzed the cells by flow cytometry and cell culture supernatants by plaque assay at 16 h , 24 h , 36 h , 48 h and 72 h after infection ( Fig . 2A and B ) . We found that LCs were infected most rapidly but that their infection frequency plateaued after 24 h . The infection kinetics were delayed for CD14+ DC and CD1c+ DCs but peaked to similar infection levels as LCs after 36 hrs . CD141+ DCs were resistant to infection throughout the entire time course ( Fig . 2A ) . Macrophage and CD14+ DCs showed similar infection kinetics but overall infection levels were lower for macrophages compared to CD14+ DCs after 36 hrs . Virus titers measured in the supernatant of infected total skin cells increased to a peak at 24 h after infection before declining . Only one out of three donors showed extended virus production until 96 h after infection ( Fig . 2B ) . To test whether and how much virus was produced by individual DC subsets , skin cells were sorted , infected , and viral RNA was extracted from the cell culture supernatants for qRT-PCR analysis . Macrophages were not viable after sorting and could not be included in this experiment . 24 h after infection , we observed the highest titers in LCs compared to the other two subsets , which showed only little ( CD1c+ ) or no increase in secreted virus ( CD14+ ) at this time point ( Fig . 2C ) . At 48 h , all infected DC subsets showed an increase in virus production , whereby LCs remained the most efficient producers . To determine the relative viral load contributed by each DC subset we first determined the relative numbers of each subset in digested whole skin ( Fig . 2D ) and then calculated their relative contribution ( Fig . 2E ) . This analysis revealed that LCs were the main contributors of viral load produced by DCs , followed by CD1c+ DCs , which were present in higher numbers than CD14+ DCs . We next assessed if DENV infection resulted in cell apoptosis . Using AnnexinV staining we found that a significantly lower proportion of infected CD1c+ DCs and LCs were apoptotic compared to their non-infected counterparts in the same tissue culture well suggesting that infection per se did not induce APC apoptosis at either 24 h after infection ( Fig . 2F ) or up to 90 h after infection ( Table S1 ) . CD14+ DCs appeared to be more susceptible to apoptosis following DENV infection ( Fig . 2F ) . Prolonged survival of infected cells might represent an effective strategy of the virus to maximize the time for production of viral progeny . In summary , we found that infected human skin DCs were capable of producing significant amounts of DENV in the absence of increased levels of apoptosis in infected cells . We next assessed whether the differential infection rates of skin APCs could be explained by variations in individual cell types' ability to take up the virus . After confirming that inactivated , fluorescently labeled virus was still able to bind to host cells ( Figure S2A ) , virus uptake was measured at 37 degrees , whereas lack of uptake at 4 degrees served as negative control ( Fig . 3A ) . 2 h after adding inactivated , fluorescently-labeled virus the highest virus uptake rate was observed in CD14+ DCs and dermal macrophages , which was in line with efficient infection . LCs , however , showed a relatively lower viral uptake activity but were still efficiently infected . This was even more surprising with regards to the expression pattern of well-described host receptors for viral binding and infection DC-SIGN ( CD209 ) and mannose receptor ( CD206 ) [21] [22] , [23] , which were absent on LCs [24] ( Figure S2B ) . The phosphatidylserine receptor Axl was recently described as alternative virus-binding receptor [25]–[27] and is expressed on LCs [28] . However , we detected negligible levels of Axl on the surface of skin DC subsets isolated by collagenase-digestion or spontaneous migration from skin explants ex vivo . DC-SIGN was expressed on CD14+ DCs but not on CD1c+ or CD141+ DCs , whereas CD206 was expressed on CD14+ and CD1c+ DCs ( Figure S2B ) . At least for dermal DCs , expression of these two receptors could therefore explain the more efficient infection of CD14+ DCs compared to CD1c+ DCs and the absence of infection in CD141+ DCs . To further study the relevance of DC-SIGN for infectivity , we first confirmed that DENV-2 D2Y98P did bind to DC-SIGN by incubating DC-SIGN-expressing U937 cells in the presence or absence of DC-SIGN blocking antibody ( Figure S2C ) . Blocking of DC-SIGN on skin APC subsets had no effect on infection rates , suggesting that other receptors utilized by DENV might be more relevant than DC-SIGN on primary skin DCs ( Fig . 3B ) . Since infection rates by the individual APC subsets might be affected by differential sensitivity to IFN , we pre-treated total skin cells with IFN-β and detected infection rates of APC subsets at 24 h by flow cytometry . Increasing concentrations of IFN-β had a significant inhibitory effect on infection in CD14+ DCs , CD1c+ DCs and macrophages but not in LCs ( Fig . 3C ) . This suggests that LCs are less sensitive to IFN , allowing the virus to replicate efficiently . Overall , infection of skin DC subsets did not strictly correlate with the expression of DC-SIGN , mannose receptor or Axl , while the extent of virus particle uptake only correlated with infection in dermal APCs , and not in LCs . These findings suggested that additional cell-inherent parameters including IFN-β susceptibility determined the observed DENV tropism for distinct skin APC subsets . To evaluate if DENV infection affected T cell stimulatory function of skin DCs , we tested the capacity of the different skin DC subsets infected with dengue virus to stimulate proliferation of allogeneic T cells ( Fig . 4A and B ) . Sorted infected DC subsets were incubated with allogeneic CD3-sorted CFSE-labeled T cells for 5 days before measurement of CD4+ and CD8+ T cell proliferation by flow cytometry ( Fig . 3A and Table S2 ) . DENV-infected CD14+ DCs were less efficient at inducing CD4+ T cell proliferation compared to their non-infected counterparts ( Fig . 4A ) . This is in keeping with previous observations of poor T cell proliferative responses when PBMCs from dengue infected patients were stimulated with PHA [29] . The defect could be restored by the addition of IL-2 or gamma-irradiated PBMCs from healthy donors , suggesting that APCs but not T cells were impaired in patients [29] . Moreover , DENV infection of monocyte-derived DC ( moDCs ) inhibited their maturation and their capacity to induce proliferation in allogeneic bulk T cells [30] , [31] , but not sorted naïve CD4+ T cells [32] , [33] . In contrast to CD14+ DCs , infection of CD1c+ DCs and LCs did not impair their capacity to induce allogeneic CD4+ T cell proliferation , showing that DENV-mediated inhibition of T cell proliferation was DC subset specific and not a direct generic effect of the virus either on DCs or T cells ( Fig . 3B ) . However , infection of DC subsets had no effect on their capacity to induce CD8+ T cell proliferation ( Table S2 ) . We next tested whether infection of skin DCs was likely to have an impact on their capacity to migrate towards the chemokine CCL19 , which is expressed in the T cell zones of LN follicles to attract CCR7-expessing migratory cells [34] . CCR7 expression levels were comparable between DC subsets exposed to DENV for 24 h and those treated with medium alone ( Fig . 4C ) , and a functional chemotaxis assay performed at different time points confirmed that DENV infection did not have an impact on the migration of skin DCs in vitro as infected cells migrated equally compared to non-infected DCs ( Fig . 4D ) . CD1c+ and LCs migrated more efficiently than CD14+ DCs in both conditions ( Fig . 4E ) . To assess the effects of DENV infection on skin DC function , we detected the effects of DENV exposure on the transcription of 184 inflammatory and immune response genes in skin DC subsets . Sorted skin DCs were infected with DENV-2 D2Y98P for 24 h and the mRNA transcripts present in cell lysates were quantified by Nanostring ( Fig . 5A ) . In these experiments , the mean infection rates were 28 . 1% for CD14+ DCs , 39 . 5% for LCs and 12 . 5% for CD1c+ DCs . Transcription of IFN-β , STAT-1 and CCL5 was significantly up-regulated in all APC subsets upon dengue virus infection . The greatest changes in expression occurred in the CD14+ DCs ( Fig . 5A ) . Of note , CD141+ DCs did not up-regulate early antiviral genes compared to the other subsets , suggesting that IFN response induction was not responsible for resistance to DENV infection . Up-regulation of IFNA1 gene expression in CD141+ cells was not statistically significant and only observed in two out of four donors studied ( Fig . 5A ) . Expression of IFN-β and CCL5 48 h after infection was confirmed by ELISA and was not seen in UV-DENV treated cells , showing that viral replication was necessary to induce innate immune gene up-regulation ( Fig . 5B ) . However , these experiments could not distinguish between gene expression in infected and non-infected cells , which might both contribute to the total gene up-regulation . Taken together , our data demonstrated that DENV virus infection of human skin DC subsets differentially affected their allogeneic T cell stimulatory capacity . Moreover , primary human skin DCs had the capacity to initiate IFN transcription upon viral infection . In addition , the rapid induction of inflammatory genes such as CCL5 could attract innate immune cells to clear local infection and possibly increase migration of DCs to draining LNs for the initiation of the adaptive response . Having identified the cell types in human skin that are able to be infected by DENV , we next wanted to understand the in vivo consequences of dengue infection on ensuing functional responses . We recently identified the functional murine homologs of human tissue DC subsets [15] , which enabled us to exploit a murine model of dengue infection to interrogate skin APC susceptibility to DENV infection , their LN migratory capacity and the contribution of recruited inflammatory myeloid cells upon DENV infection . As wild-type mice are not susceptible to DENV infection [35] , we used interferon-α/β-receptor knock-out ( IFNAR ) mice , which show disease symptoms and clinical parameters comparable to dengue patients following DENV infection [36] . Mice were infected intra-dermally with 106 pfu of DENV-2 in a volume of 10 ul into each ear . After two and four days mice were sacrificed and we prepared single-cell suspensions from their ears for flow cytometry analysis of infected cell populations staining positively for DENV E protein . The gating strategy ( Figure S3A ) allowed us to differentiate between dermal CD11b+ DCs ( homolog of human CD1c+ DCs ) , CD11b− DCs , CD103+ DCs ( homolog of human CD141+ DCs ) , LCs , MHC class II ( IAIE ) hi Ly6C+ monocyte-derived cells and MHC class II ( IAIE ) − Ly6C+ inflammatory monocytes [37] , [38] . CD11b− and CD11b+ dermal DCs were frequently infected , reaching infection levels of approximately 20% and 50% respectively by day 4 post-infection ( Fig . 6A and B ) . In contrast , by day 4 , LCs were not highly infected and similarly CD103+ DCs showed a low level of infection in the region of 10% of cells by day 2 ( Fig . 6B ) . In addition to skin-resident DCs , we found high infection rates in infiltrating Ly6C+ cells in the skin two days after infection ( Fig . 6A ) , with a marked increase in infection particularly in the Ly6C+IAIE+ population on day four after infection ( Fig . 6B ) . In contrast to ex vivo human skin cells , we also identified a substantial population of infected CD45− cells in the dermis , but not in the epidermis ( Figure S3B ) . Quantification of total cell numbers 2 and 4 days after infection showed a decrease , although not significant , of CD11b− DCs , CD11b+ DCs and CD103+ DCs , two days after infection . This was likely due to cell death rather than LN migration as the increase in numbers of migrated cells in draining LNs was already observed at day 2 after infection ( see following paragraph ) . In contrast to dermal DCs , LC numbers remained constant ( Fig . 6C ) . The murine in vivo model also permitted analysis of cells recruited into skin upon intradermal inoculation of DENV . We observed a more than ten-fold increase in the number of inflammatory monocytes ( Ly6C+IAIE− ) and monocyte-derived cells ( Ly6C+IAIE+ ) in the ears of mice within two days of infection . This rise was followed by a rapid decline four days after infection , which may be due to cell death or due to down-regulation of Ly6C expression on activated monocyte-derived cells . Ly6C− monocyte-derived cells could not be distinguished from CD11b+ DCs and it was difficult to assess whether there was a relatively smaller decline in this population due to possible parallel effects of cell death and new formation of monocyte-derived cells ( Fig . 6C ) . Taken together , functionally-equivalent dermal DC subsets appeared to be infected in both humans and mice , with the exception of CD103+ cells , which were infected in mouse skin , although at much reduced numbers compared to other subsets , whereas their counterpart in human skin were not infected in our ex vivo experiments . In addition , in vivo experiments revealed a massive infiltration of Ly6C+ monocyte-derived cells , identifying these cells as potentially important infection targets during natural infection . It was important to know which cells had the capacity to migrate to draining LNs for the initiation of adaptive immune responses and the ensuing immune memory , and whether those cells carried infectious DENV with them . Cell suspensions were made from ear-draining LNs of infected mice at days 2 and 4 , and analyzed by flow cytometry for DENV E protein , with immigrant and resident DC discriminated based on CD11c and MHC Class II ( IAIE ) expression [39] , [40] ( Fig . 7A–C ) . Amongst DCs migrated from the skin , CD11b− and CD11b+ DCs , but not LCs were infected . Despite the low CD103+ DC infection rate compared to CD11b+ DC in the skin , more infected CD103+ DCs than CD11b+ DCs were observed in the draining LN at day 4 ( Fig . 7C ) . The relative abundance of CD103+ DCs in the draining LN compared to other infected subsets migrating from skin , suggests a significant role for this subset in T cell activation at later time points of infection . The LN-resident counterparts of CD103+ cells , CD8+ DCs , were not infected at the time points tested , suggesting that virus is transported from the skin via DCs and that little virus reaches the LN directly via the lymphatics to infect LN-resident DCs or these cells were not susceptible to infection at this stage ( Fig . 7B and C ) . Similarly , LN-resident CD11b+ DCs were also not infected . However , the absolute number of ( non-infected ) LN-resident CD11b+ DCs was significantly greater in infected compared to non-infected mice ( Fig . 7D ) . In summary , infection of LN-resident DCs was negligible in contrast to active infection of skin APCs within the first 4 days after intradermal DENV delivery . This finding suggested that dermal DCs migrating from the skin to draining LNs were efficient carriers of infectious DENV and implicates them as likely initiators of systemic immunity . Our results further indicated that CD11b+ dermal DCs ( the equivalents of human CD1c+ DCs ) might be important to trigger early adaptive anti-DENV T cell responses .
The aim of this study was to characterize the cellular targets of DENV infection in human skin , and the consequences of APC infection on systemic infection and the induction of a protective immune response ( see model , Fig . 8 ) . Previous studies have focused on the role of LCs in DENV infection , but dermal DCs were not evaluated [41] . DENV is most likely injected into the dermis by its mosquito vector [42] and we provide evidence that human dermal APCs can also be efficiently infected by DENV . In fact , when DENV is injected intradermally ex vivo or in mice , LCs are not infected efficiently and the functional relevance of natural LC infection therefore remains unclear . Virus might come into contact with LCs during the process of probing even though the mosquito's proboscis bypasses the epidermis and LCs located there . Alternatively , LCs that migrate through the dermis towards draining LNs might be infected en route , although the number of spontaneously migrating LCs in healthy skin is very small [43] . The suspension cell infection model cannot solve this question and further experiments , ideally with infected mosquitoes injecting the virus into the skin of mice or other animal models , will be required to validate our findings . Interestingly , we found that DC-SIGN expression did not correlate with infection and blocking of the receptor did not reduce the rate of infection in cells expressing DC-SIGN . Monocyte-derived DCs ( moDCs ) [44] , which represent an easily accessible and hence useful model to study DCs in vitro , express high levels of DC-SIGN . The majority of primary DC subsets found in blood , skin and lymph nodes , however , does not express DC-SIGN when analyzed ex vivo [45] , [46] . It was previously shown that DC-SIGN facilitates attachment to the cell rather than mediating viral endocytosis per se and that a potentially unknown bona fide receptor is required for viral entry [47] . We therefore speculate that DC-SIGN is one of several receptors expressed on primary DCs that facilitate attachment and viral entry [48] . In mice , CD11b+ DCs , which are the functional homologs of human CD1c+ DCs , had the highest infection rate . In addition , recruited inflammatory Ly6C+ monocyte-derived cells into skin were also efficiently infected ( Figure S3 , population 6 ) . Based on the findings in mice , we speculate that human blood monocytes infiltrate the skin upon infection , similar to Ly6C+ mouse monocytes , and represent an additional target population for the virus . Addressing this question in humans in vivo is a challenge , as skin biopsies from patients from the site of the mosquito bite would have to be analyzed . Human steady state CD14+ dermal DCs were efficiently infected . This subset expressed low levels of CCR7 ( Fig . 4C ) and is unlikely to migrate efficiently to draining LNs ( Fig . 4E and [12] ) . However , CD14+ cells are efficient activators of memory T cells [12] suggesting a role in local tissue responses , particularly during secondary infection when DENV-specific T cells may be present in the skin . Alternatively , infection of CD14+ DCs could affect their capacity to induce regulatory CD4 T cells in the skin , a function that has been associated with CD14+ DCs and not with CD1c+ DCs [5] , [14] . We were unable to establish in the in vivo model if infiltrating Ly6C+IAIE+ cells migrated from the skin to the draining LNs as Ly6C could be downregulated during LN migration as shown for a West Nile virus murine infection model [49] . Despite the obvious similarities between human and mouse skin infection targets there were also notable differences: firstly , we observed a substantial population of CD45− infected cells in mouse , but not human , skin . This is challenging to interpret , but the lack of the IFNα/β receptor in these mice could have affected the susceptibility of non-hematopoietic cells . We showed recently that the absence of the IFNα/β receptor on CD11c+ and LysM+ expressing cells alone was sufficient to replicate the DENV-susceptibility phenotype of IFNAR−/− mice with regards to viremia and survival [35] , though this does not directly exclude the possibility of some alterations to viral tropism within the model . The second difference between human and mouse was that human CD141+ DCs remained uninfected up to 72 h after infection ex vivo , whereas CD103+ DCs in mice were infected at day four after infection . It could be that CD141+ DCs may become susceptible at later time points after infection ex vivo or that CD141+ DCs are infected in the context of a natural infection . It remains to be addressed whether the reason for this discrepancy relates to species-specific differences in the antiviral response . For human CD141+ DCs , infection-induced up-regulation of IFN-β and STAT-1 gene expression was low compared to CD14+ DCs and CD1c+ DCs ( Fig . 4 ) , which might also reflect the lack of infection of the cells at these time points . The induction of transcription of the monocyte-attracting cytokine CCL5 in human cells fits nicely with our observation of infiltrating monocyte and monocyte-derived cells in mice , making it likely that monocyte-derived cells are similarly attracted to the site of infection in humans [50] . IFN signatures and CCL5 were previously found to be up-regulated in microarrays of dengue patients' PBMCs [51] , [52] and in the serum [50] , whereby higher expression seemed to be associated with less severe disease . We demonstrate here that human skin DCs are likely to be important targets for DENV infection in vivo . The observations in mice suggest that skin dermal DCs were also likely to transport infectious virus to draining LNs , providing a shuttle for the virus to potentially establish further sites of infection , and to efficiently activate a systemic immune response . Our data suggest that intra-dermal or gene-gun inoculation of live-attenuated dengue vaccine candidates would likely target the most physiologically relevant DC populations within the dermis and thereby potentially stimulate the efficient establishment of a systemic immune response .
Healthy human skin tissue was obtained from mastectomies or abdominoplastic surgery . The studies were approved by the respective institutional review boards ( National Health Group Domain Specific Review Board ( NHG DSRB 2012/00928 ) and Singhealth Centralized Institutional Review Board ( CIRB 2011/327/E ) , respectively ) and patients gave written informed consent . All skin samples were processed on the day of surgery . Blood from anonymous healthy human donors was received from the blood bank at the National University Hospital of Singapore and blood donors gave written informed consent . The study was exempted from full IRB review by the Institutional Review Board of the National University of Singapore ( NUS-IRB ) since anonymous samples were used . Mouse experiments were conducted according to the rules and guidelines of the Agri-Food and Veterinary Authority ( AVA ) and the National Advisory Committee for Laboratory Animal Research ( NACLAR ) , Singapore . The experiments were reviewed and approved by the Institutional Review Board of the Biological Resource Center , Singapore ( IACUC protocols 100566 and 120801 ) . IFN-α/β receptor-deficient mice ( IFNAR−/− ) , on a C57BL/6 background , were infected with 2×106 pfu of D2Y98P via the intradermal ( i . d . ) route in the ears using a 33-gauge needle and a microsyringe ( Nanofil ) . Naïve mice served as control . For isolation of human skin cells 300 µm dermatome sections were incubated in RPMI+10%FCS ( BioWest ) containing 0 . 8 mg/ml collagenase ( Type IV , Worthington-Biochemical ) and 0 . 05 mg/ml DNase I ( Roche ) for 12 h . For nanostring analysis and T cell proliferation assay skin was treated with 1 mg/ml dispase ( Invitrogen ) to separate epidermis and dermis . Dermal DCs were sorted by fluorescence-activated cell sorting ( FACS ) , epidermal LCs were isolated using CD1a microbeads ( Miltenyi Biotec ) and a magnet ( Stemcell techonologies ) with a purity of >90% . For isolation of mouse skin cells , mice were sacrificed and ears were cut off at the base . Ear skin was split into dorsal and ventral halves and incubated in RPMI+10%FCS containing 1 mg/ml dispase ( Invitrogen ) for 2 h at 37deg . Epidermis and dermis were separated and digested in 0 . 2 mg/ml collagenase ( Type IV , Sigma ) for 2 h at 37deg before passing them through a 70 um filter to obtain a single cell suspension . Mouse skin-draining auricular lymph nodes were isolated , incubated in medium+0 . 2 mg/ml collagenase for 30 min and passed through 70 um filter . BHK-21 and C6/36 cells were purchased from the American Type Culture Collection . For infection experiments the following strains of dengue virus were used: DENV-1 – 08K3126 , DENV-2 - TSV-01 or D2Y98P , DENV-3 – VN32/96 , and DENV-4 – 2641Y08 . All strains are patient isolates that have been passaged in C6/36 mosquito cells for 5–20 passages . D2Y98P used here was plaque-purified after passage 20 and derived from an infectious clone [19] . The enhanced viral RNA synthesis capacity of D2Y98P was mapped to a natural mutation in NS4b . The mutation had no effect on the IFN-inhibiting capacity of the virus [19] . All strains viruses used in the experiments were produced in the C6/36 mosquito cell line . For phagocytosis assays , DENV-3 - VN32/96 was purified with density gradient isolation ( Opti-Prep , Sigma ) according to manufacturer's protocol . Viral particles were labeled with Alexa-647 fluorescent dye using a protein labeling kit ( Molecular Probes ) and the excess dye was removed with Amicon protein purification tubes ( Millipore ) according to manufacturer's protocol . The virus was then inactivated in 2 mM DEPC/PBS-T for 15 min at RT [53] . Flow cytometry was performed on an LSRII , FACSCanto , FACS was performed using a FACSAriaII ( all Becton Dickinson [BD] ) . Software analysis was performed with FlowJo ( TreeStar ) . The following reagents for labeling of human cells were used: Carboxyfluorescein succinimidyl ester ( CFSE ) , fixable live/dead blue dye ( Life Science Technologies ) , anti-CD3 ( UCHT1 ) , anti-CD4 ( RPA-T4 ) , anti-CD8 ( RPA-T8 ) , anti-CD1a ( HI149 ) , anti-CD209 ( 9E9A8 ) , anti-CD206 ( 15-2 ) ( all from Biolegend ) , anti-CD11c ( B-ly6 ) , AnnexinV Detection Kit , anti-CD45 ( HI30 ) , anti-HLA-DR ( L243 ) ( all from BD Biosciences ) , anti-CD141 PE ( AD5-14H12 ) ( Miltenyi ) , anti-CD14 ( RMO52 ) ( Beckman Coulter ) , anti-CCR7 ( 3D12 ) ( eBioscience ) , anti-Axl ( MAB154 ) ( R&D Systems ) and anti-E protein ( 4G2 ) ( ATCC ) . The following antibodies were purchased from Biolegend to label mouse cells: anti-CD45 ( 30-F11 ) , anti-IAIE ( M5/114 . 15 . 2 ) , anti-CD11b ( M1/70 ) , anti-CD11c ( N418 ) , anti-CD326 ( EpCAM ) ( G8 . 8 ) , anti-CD103 ( 2E7 ) , anti-Ly6C ( HK1 . 4 ) . Sorted DCs were infected for 2 h and immediately co-cultured with allogeneic CD3+ flow-sorted CFSE-labeled T cells from healthy blood donors in a ratio of 1∶10 in 96-well U-bottom plates for 5 days before proliferation was determined by CFSE dilution . Cell migration was assayed in chemotaxis microchamber plates ( Neuroprobe ) containing a membrane with 5 µm pores . Briefly , medium alone or containing recombinant human CCL19 ( 20 ng/ml , R&D Systems ) was added to the lower chamber . The membrane was placed on top and a cell droplet ( containing approximately 250 , 000 cells ) was pipetted on top of the membrane . Plates were incubated for 2 h at 37 degrees C and relative cell numbers of migrated cells were determined using a CellTiter Glo luminescent cell viability assay , read on a GloMax-96 microplate luminometer ( both from Promega ) . Nanostring analysis and initial data processing was performed in the nCounter system according to manufacturer's instructions . The human inflammation gene cartridge ( GXA-IN1 ) was used , and based on the data PGK1 , TUBB and GAPDH were used as housekeeping controls . Differential expression analysis was determined with a 2-way ANOVA using celltype and infection status as factors in R v2 . 15 . 2/Bioconductor . Multiple testing correction was performed using the method of Benjamini and Hochberg . Heat maps were generated using the logarithmically transformed fold changes of averaged normalized counts for each cell population using the non-infected samples as the reference . Visualization of the data and test results were done using TIBCO Spotfire . NCBI accession numbers of all genes are listed in Table S3 . Levels of CCL5 and IFNβ in skin cell supernatants were measured by enzyme-linked immunosorbent assay ( ELISA ) ( both R&D Systems ) following the manufacturer's instructions . Virus titer cell culture supernatant was determined by plaque-forming assay using BHK-21 cells as described elsewhere [35] . Briefly , supernatant of infected cells was diluted 10-fold on BHK-21 cells . After 1 h , medium was exchanged for 0 . 8% methylcellulose in RPMI/10%FCS and plates were incubated for 4 days . Plaque counts were used to calculate viral titer in plaque forming units per ml . Viral RNA was extracted from cell supernatants using a viral RNA extraction ( Roche ) according to the manufacturer's protocol and subsequently quantified by real-time qRT-PCR using primers and methods reported previously [54] . Forward primer ACACCACAGAGTTCCATCACAGA , reverse primer CATCTCATTGAAGTCNAGGCC , probe CGATGGARTGCTCTC . Binding experiments were performed with U937 cells stably expressing DC-SIGN [54] . Virus was incubated with the cells for 1 h at 4°C and cells were subsequently washed with serum-free medium . Non-fluorescently labelled virus was detected with 4G2-A647 anti-E antibody . For DC-SIGN blocking experiments cells were pre-incubated with 20 ug/ml anti-DC-SIGN mAb ( clone 120507 ) or a matched isotype controls ( clone 133303 ) ( R&D Systems ) for 1 h at 37°C . | Dengue virus ( DENV ) is transmitted by mosquitoes with skin as point of entry for the virus . Here , we investigated DENV infection in primary human skin cells and their initial immune response . Using skin from normal human donors for infection with DENV in vitro we identified antigen-presenting cells ( APCs ) as main targets of DENV . Further analysis showed that only distinct subsets of dendritic cells ( DCs ) and macrophages were infected and efficiently produced viral progeny . Langerhans cells were most susceptible to infection despite lacking DC-SIGN , a previously described DENV receptor . Infection of the other DC subsets and macrophages was also independent of DC-SIGN expression . Genes of the interferon pathway and CCL5 , a chemokine attracting immune cells to sites of inflammation , were highly up-regulated in the infected DC subsets . Using a mouse infection model , we showed that murine dermal DCs were also susceptible to DENV and migrated to draining lymph nodes . At the same time infiltrating monocytes differentiated into monocyte-derived cells at the site of infection and became an additional target for DENV in vivo . These data demonstrate that DENV differentially infects and activates primary human skin APCs and that infected cell types individually contribute to inflammation and the adaptive response . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"interdigitating",
"cells",
"classical",
"immunology",
"langerhans",
"cells",
"dermatology",
"dendritic",
"cells",
"medicine",
"and",
"health",
"sciences",
"skin",
"infections",
"immune",
"cells",
"lymphocyte",
"activation",
"cell",
"biology",
"animal",
"cells",
"antigen-presenting",
"cells",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"immunology",
"immune",
"response"
] | 2014 | Selective Susceptibility of Human Skin Antigen Presenting Cells to Productive Dengue Virus Infection |
The processes regulating peripheral naive T-cell numbers and clonal diversity remain poorly understood . Conceptually , homeostatic mechanisms must fall into the broad categories of neutral ( simple random birth–death models ) , competition ( regulation of cell numbers through quorum-sensing , perhaps via limiting shared resources ) , adaptation ( involving cell-intrinsic changes in homeostatic fitness , defined as net growth rate over time ) , or selection ( involving the loss or outgrowth of cell populations deriving from intercellular variation in fitness ) . There may also be stably maintained heterogeneity within the naive T-cell pool . To distinguish between these mechanisms , we confront very general models of these processes with an array of experimental data , both new and published . While reduced competition for homeostatic stimuli may impact cell survival or proliferation in neonates or under moderate to severe lymphopenia , we show that the only mechanism capable of explaining multiple , independent experimental studies of naive CD4+ and CD8+ T-cell homeostasis in mice from young adulthood into old age is one of adaptation , in which cells act independently and accrue a survival or proliferative advantage continuously with their post-thymic age . However , aged naive T cells may also be functionally impaired , and so the accumulation of older cells via ‘conditioning through experience’ may contribute to reduced immune responsiveness in the elderly .
Naive T cells accumulate in the periphery rapidly from birth , but their numbers decline gradually from puberty onwards in both mice and humans due to the slow involution of thymus and associated decline in the export of new cells [1 , 2] . Despite substantial knowledge of the qualitative nature of the cues involved in their survival and proliferative renewal—which include signals through the T-cell receptor ( TCR ) and from cytokines—we have a relatively limited quantitative understanding of how the total numbers and receptor diversity of naive T cells are determined throughout life . The consensus in the field has been that the population dynamics of naive T cells are influenced by intra- and/or interclonal competition for limiting homeostatic cues , largely motivated by observations that homeostatic proliferation and cell longevity increase under severely lymphopenic conditions [3–7] . In support of this hypothesis , mathematical models of resource competition—in which all cells compete for a limiting , ‘public’ supply of homeostatic stimuli—have successfully described naive T-cell population dynamics in lymphoreplete and partially lymphopenic settings [8 , 9] . However , multiple observations indicate that these models have limited explanatory power . The extent to which resource competition , or any similar quorum-sensing mechanism , influences cell lifetimes or division rates under replete conditions is unclear [10–12] , and resource competition alone is unable to explain the kinetics of replacement of old naive T cells by new cells exported from the thymus in healthy mice [13] . There is also evidence that naive T cells’ homeostatic fitness , defined as the difference between their rates of division and loss , may vary with host or cell age . Naive TCR transgenic T cells from aged mice persist longer than the same cells from young mice following transfer , and naive T cells are lost more slowly following thymectomy in old mice than in young mice [14] . There are at least two mechanisms that may generate heterogeneity in homeostatic fitness and potentially explain these observations . One is a process of adaptation , in which cells accumulate changes , possibly in response to microenvironmental signals , that improve survival or the ability to self-renew through division the longer they survive [13 , 15] . Such changes might , for example , reflect the continued maturation of recent thymic emigrants ( RTEs ) in the periphery [16 , 17] . Another mechanism is a process of selection acting upon cell populations exhibiting a distribution of stable , cell-intrinsic rates of homeostatic division or loss [18–20] . This variation in fitness might derive from clone-specific differences in TCR affinity for self-peptide–MHC ligands [21] or potentially levels of expression of the interleukin 7 ( IL-7 ) receptor [22] , although to our knowledge the impact of the latter is manifest only in lymphopenia [23] . Modelling has quantified the extent to which heterogeneity in naive cells’ capacity to survive can select for intrinsically long-lived cells over time [13 , 19 , 24] . Crucially , however , adaptive and selective mechanisms are difficult to distinguish directly using standard cross-sectional studies of T-cell dynamics in mice because they can give rise to qualitatively similar distributions of fitnesses at the cell population level . Mathematical models are appropriate tools for describing cell population dynamics , and when combined with experimental data , they can boost our ability to discriminate between candidate immunological mechanisms . Furthermore , a model’s worth can be assessed both by its ability to explain a particular dataset ( its goodness of fit ) as well as its ability to successfully describe multiple independent experiments ( its robustness ) . With this philosophy in mind , we confronted models of different homeostatic mechanisms with a diverse set of data , generated by ourselves and from other laboratories . Specifically , we compared different models’ abilities to explain ( i ) naive T-cell dynamics in healthy , replete mice and following thymectomy , ( ii ) the dynamics of infiltration of new naive cells from the thymus into an intact peripheral compartment , and ( iii ) cotransfers of cells of different ages into the same host . Together , these data allowed us to rule out purely homogeneous models in which all naive T cells compete equally for homeostatic stimuli . Furthermore , we found that a model of cellular adaptation , in which a cell’s fitness increases with its post-thymic age , was uniquely able to explain all three disparate datasets . While adaptation , selection , and quorum-sensing are not mutually exclusive and may operate in combination , we argue that , under normal physiological circumstances , adaptation is the dominant force shaping naive T-cell population dynamics because in isolation it provides a parsimonious explanation of diverse experimental data .
We explored five types of behaviour underlying naive T-cell population dynamics , which we describe conceptually and mathematically here and illustrate in Fig 1 . Full details of the formulation and solution of these models as well as the procedures for fitting them to experimental data are given in S1 Text . We use the standard definition of cell ‘fitness’ as a measure of reproductive success , or a net growth rate—i . e . , the propensity of a cell for division minus its propensity for loss . Therefore , ‘homeostatic fitness’ is an absolute , not a relative , measure of the ability of a T cell and its progeny to persist within the naive T-cell compartment over time . The protocol used to generate busulfan chimeras is described in detail elsewhere [13 , 26] . The following monoclonal antibodies and cell dyes were used: CD45 . 1 FITC , CD45 . 2 AlexaFluor700 , CD45 . 2 FITC , TCR-beta APC , CD4 PerCP-eFluor710 , CD44 APC-eFluor780 , CD25 PE-Cy7 , L-selectin eFluor450 , CD122 biotin ( all eBioscience ) ; CD8 Pacific Orange , streptavidin PE-TexasRed , LIVE/DEAD blue ( all Invitrogen ) ; and CD45 . 1 Brilliant Violet 650 , CD4 Brilliant Violet 711 , and TCR-beta PerCP-Cy5 . 5 ( all BioLegend ) . Samples were acquired on LSR-II , LSRFortessa , or Fortessa X20 flow cytometers ( BD ) , and analysis was performed with FlowJo software ( Treestar ) . Wild-type ( WT ) CD45 . 1 and CD45 . 2 mice were bred and maintained in conventional pathogen-free colonies at either the National Institute for Medical Research ( London , United Kingdom ) or at the Royal Free Campus of University College London . All experiments were performed in accordance with UK Home Office regulations , project license number PPL70-8310 .
The first test of the models was to confront them with data from a study that measured the numbers of naive ( CD62LhiCD44lo ) CD4 and CD8 T lymphocytes recovered from spleen and lymph nodes in euthymic ( WT ) and thymectomised ( Tx ) C57BL/6 mice , from soon after birth up to 65 wk of age [8] . Thymectomy was performed at 7 wk of age . We found that the neutral model of random division and loss at constant rates throughout life reproduces the long-term kinetics of both naive CD4 and CD8 T cells in WT mice remarkably well ( Fig 2A , solid lines ) . These changes in naive T-cell numbers in mice under normal conditions can therefore be explained broadly without invoking any compensatory changes in the division or cell longevity with pool size or waning thymic output . However , the neutral model clearly fails to capture the kinetics with which naive CD4 and CD8 T-cell numbers fall following thymectomy ( Fig 2A , dashed lines ) . In the original study of den Braber et al . [8] , a resource competition model was invoked to explain the improved net survival of cells in the absence of thymic export . We replicated their fits using the same model , confirming that a density-dependent net rate of loss provides a substantially better description of both naive CD4 and CD8 kinetics—in WT and Tx mice simultaneously from 7 wk of age—than the neutral model ( Fig 2B , orange lines; ΔAIC = 91 and 115 , respectively; ΔAIC refers to differences in the Akaike Information Criterion [27] ) . Given the evidence for heterogeneity within the naive compartments , we then assessed the abilities of the adaptation , selection , and incumbent models to explain naive T-cell kinetics from the age at thymectomy ( 7 wk ) onwards , comparing them to the density-dependent model . The adaptation , selection , and density-dependent models produced visually similar fits ( Fig 2B ) , but the selection model received the strongest statistical support ( Table 1 , ΔAIC = 12 ) . In contrast , the incumbent model gave the poorest fit ( ΔAIC = 29 ) . Therefore , naive CD4 and CD8 T-cell kinetics from 7 wk of age onwards in euthymic and Tx mice—during which cell numbers vary by a factor of approximately 10—can be explained well without invoking any density-dependent processes . The data included observations from mice soon after birth , when mice might be considered lymphopenic [28] , and so any density-dependent effects might be more apparent . Assuming that thymocyte numbers from birth into old age continue to reflect thymic output with the same constant of proportionality , we fitted the density-dependent , adaptation , and selection models to the entire time courses ( Fig 2C ) . It was not possible to fit the incumbent model to these data because the kinetic by which any such cells are established is unknown . Here , we found greater support for the density-dependent model over the selection and adaptation models ( ΔAIC = 3 . 6 and 9 . 2 , respectively , for CD4; 66 and 71 , respectively , for CD8 ) , presumably due to its improved description of the additional data describing the growth of naive T-cell numbers in young mice . To examine this more closely , we fitted the neutral and density-dependent models to the data up to 7 wk only , a period during which naive T-cell numbers increase rapidly . In this early window , we found that the neutral and density-dependent models had essentially equivalent statistical support ( ΔAIC = 0 . 43 in favour of the neutral model for CD4; 0 . 02 for CD8 ) . Therefore , our analysis provides no strong evidence that lymphopenia-induced proliferation through reduced competition for homeostatic stimuli is a dominant factor in establishing the naive CD4 and CD8 T-cell pools early in life . Previously , we used a bone marrow chimeric system to study the homeostatic dynamics of T cells [13 , 29] . CD45 . 1 C57BL/6 mice were treated with optimised doses of the transplant-conditioning drug busulfan , specifically ablating haematopoetic stem cells ( HSCs ) but leaving thymic and peripheral lymphocyte compartments intact . Following congenic ( CD45 . 2 ) bone marrow transplant ( BMT ) , the appearance of donor-derived cells serves as a proxy for quantifying bone marrow output following BMT ( Fig 3A ) . Therefore , the combination of the kinetics of the compartment sizes and the infiltration of donor-derived cells ( Fig 3B and 3C ) allows us to quantify rates of turnover and the rules of replacement . The replacement of naive ( CD62Lhi CD44lo CD25− ) CD4 and CD8 T-cell compartments post-BMT stabilise at 80% to 90% of the levels in the thymus [13] . This shortfall is not an artefact of the experimental manipulation because both γδ T-cell and B cell populations in the same mice undergo near 100% replacement post-BMT ( Hogan , Rane , Yates , and Seddon , unpublished observations ) . The incomplete replacement of host naive cells could not be explained by thymic involution nor by any homogeneous models in which all cells obey the same kinetics [13] and which include the neutral and density-dependent models considered here . Rather , we found support for a model in which slowly self-renewing , ‘incumbent’ subpopulations of naive CD4 and CD8 T cells established early in life resist displacement by naive T cells generated later in life . In that study , we found marginally lower support for an adaptation model identical to the one described here . However , because the observations came from mice that underwent BMT at identical ages , our ability to dissociate any effects of host age and cell age was relatively limited . To address this problem , we performed additional experiments analysing the reconstitution dynamics of busulfan chimeras in hosts whose age at BMT ranged from 6 wk to 6 mo ( Materials and methods; data are provided in S2 Data ) . This expanded range provided a richer dataset , requiring models to be able to account for any impact of the ageing host environment as well as of cell age . Models were fitted simultaneously to the time courses of total naive T-cell numbers and the normalised peripheral chimerism ( i . e . , the donor proportion in the peripheral naive T-cell population divided by the stable donor proportion in all thymic compartments ) . First , as a consistency check , we fitted a density-dependent model and found that it described the data poorly ( Table 1 and S1 Fig ) , supporting our original conclusion that homogeneous models are unable to explain the replacement kinetics in this system . In contrast , the incumbent , adaptation , and selection models yielded good and visually similar fits to both total cell numbers ( Fig 3B ) and peripheral chimerism ( Fig 3C ) for both naive CD4 and CD8 T cells . Within this group of candidates , the incumbent model had the strongest support , and the selection model had the weakest ( Table 1 ) . However , recall that the incumbent model described the long-term kinetics of T-cell numbers in WT mice and mice thymectomised at 7 wk of age relatively poorly ( Table 1 ) . Therefore , by assessing both quality of fit and model robustness , we find ( i ) relatively little support for the incumbent model and ( ii ) equivocal support for the adaptation and selection models , which—to opposing degrees—are able to explain both the WT/Tx data and the replacement kinetics in the busulfan chimeras . The adaptation , selection , and incumbent models all exhibit differential growth or loss of populations due to population-level variability in homeostatic fitness . This variability may derive from progressive conditioning with time spent in the periphery ( adaptation ) , be generated during thymic development ( selection ) , or be rooted in differences in fitness between the cells that populate the empty neonatal enivironment and those that enter the replete naive compartment later in life ( the incumbent model ) . In all three scenarios , more proliferative or longer-lived naive T cells accumulate , but the mechanisms are difficult to directly distinguish experimentally without tracking the fate and fitness of individual cells over long timescales . Aiming to differentiate the mechanisms further , we turned to data from a study in which naive CD4 T cells from donor C57BL/6 mice of different ages were transferred into young recipients ( ref . [14] and S3 Data ) . In that study , polyclonal naive CD4 T cells from older hosts clearly exhibited a survival advantage over those from younger hosts over the 20 d following adoptive transfer ( Fig 4A ) . To test the abilities of the adaptation , selection , and incumbent models to explain these data , we simulated the experiment using these models to predict the kinetics of cohorts of naive CD4 T cells sampled from hosts aged 2 and 20 mo ( Fig 4A ) . We performed these simulations with two sets of parameters—those estimated using the data in den Braber et al . [8] and those from the busulfan chimeras . To compare these predictions with the data , we estimated only a single quantity: the number of cells recovered immediately following the transfer . The selection model failed to reproduce the kinetics of donor cell loss . In contrast , the adaptation and incumbent models predicted these kinetics remarkably accurately for both sets of parameters ( Fig 4A ) , with adaptation receiving the strongest statistical support ( Table 2 ) . In another experiment reported in Tsukamoto et al . [14] , TCR transgenic ( AND ) naive CD4 T cells from C57BL/10 donors of ages 2 , 6 , 12 , and 18 mo were transferred into young hosts , revealing a progressive increase in cell survival with donor age . A key observation was that the loss rate of cells derived from 12-mo-old donors was significantly greater than of cells from 18-mo-old donors . The authors argued that because thymic output drops to low levels by 6 mo of age in mice [2 , 30] , any selection for fitter cells ought to be complete by 1 year of age , and so there should be little difference in the rates of loss of cells from donors older than this . They therefore favoured adaptation over selection as an explanation for these data . To examine this argument quantitatively , we simulated the experiment using the adaptation , incumbent , and selection models , in each case using the parameters from fits to both the data from den Braber et al . ( Fig 2 ) and our busulfan chimeras ( Fig 3 ) . Notably , Tsukamoto et al . consistently observed that the AND transgenic T cells were lost more rapidly than polyclonal T cells from age-matched mice , and so our model predictions—derived from polyclonal T-cell dynamics—overestimated the observed rates of survival of AND T cells over 15 d post-transfer ( reproduced in Fig 4B , left panel ) . However , all three models allowed us to predict the trend in 15-d survival with donor age ( Fig 4B ) . We confirmed the argument in Tsukamoto et al . [14] that adaptation , but not selection , can generate a consistent increase in the recovery of cells from donors up to 18 mo old . The incumbent model also fails to capture this trend , predicting a similar fold loss for 2- , 6- , and 12- mo-old donors . This is because the proportion of the pool occupied by the more persistent incumbent cells is predicted to increase only slowly with host age [13] . We conclude that adaptation is the strongest candidate model of naive T-cell homeostasis that we considered , in terms of both the quality of its descriptions of the data and the diversity of datasets it could describe ( Table 3 ) . Another test of a model is the extent to which its parameters agree across fits to multiple datasets . We therefore compared the parameters that we derived from the data from den Braber et al . [8] and the busulfan chimeras . We found consistent estimates of the naive CD4 or CD8 T-cell pool sizes and rates of thymic output at 7 wk of age ( Table 4 ) , but estimated that fitness increases more rapidly with cell age in the WT/Tx mice studied in ref . [8] than in the busulfan chimeras . This difference is more evident for naive CD8 T cells , largely because den Braber et al . saw a rapid loss of these cells immediately post thymectomy , followed by a slower decline ( Fig 2 , right hand column ) , consistent with a more bimodal distribution of cell lifetimes with cell age than the smoother , exponential one we inferred from the busulfan chimera data . Correspondingly , the two datasets were best described with different functional forms of the declining loss rate λ ( a ) . Nevertheless , the two models give very similar predictions of the age structure of the naive CD4 T-cell pool for different host ages ( Fig 5 ) . Both show a high preponderance of younger cells—consistent with the consensus that there is a dominant role for thymic export in maintaining naive T-cell numbers in mice [8 , 13]—and a clear accumulation of older cells over time . Newly developed T cells ( RTEs ) take time to reach full functional competence in the periphery [16 , 17] , and it seems plausible that this transition may be accompanied by changes in their capacity to survive or self-renew . It is therefore possible that the adaptation process we model , in which homeostatic fitness changes with post-thymic age , may reflect the process of RTE maturation . In our analyses of the adaptation model , we explored net loss rates λ ( a ) that decline smoothly with cell age , either exponentially or with a sigmoid form ( S1 Text ) . While RTEs remain difficult to define precisely and thus the kinetics of their maturation remain unclear , we explored an alternative model of adaptation in which RTEs mature abruptly after a fixed time aM in the periphery and have a distinct loss rate ( λRTE ) from that of mature naive cells ( λM ) . As expected , we found that this stepwise , or ‘conveyor belt’ , RTE maturation model predicted that mature naive cells are lost at a lower net rate than RTEs . However , the model yielded inferior fits to naive CD4 and CD8 counts in WT and Tx mice ( ΔAIC = 83 for CD4 and 64 for CD8 cells ) and fitted the data from the busulfan chimeras poorly ( not shown ) . Nevertheless , the model predicted maturation times aM of 6 and 2 wk for CD4 and CD8 T cells , respectively , which are in approximate agreement with the estimate of 3 wk for CD4 and CD8 RTEs combined in the studies by Berzins et al . [31 , 32] and closely comparable to a more recent estimate of the expected time to maturation of CD4 RTEs [33] . We conclude that there is little support for a deterministic , conveyor belt model in which maturation of RTEs is accompanied by a rapid increase in homeostatic fitness . However , because the true kinetics of RTE maturation are unknown and may be more extended in duration , we cannot rule it out as the biological underpinning of any more gradual increase in fitness with cell age . Quorum-sensing models such as the one expressed in Eq 2 predict a unique set point or carrying capacity for the naive T-cell pool size for any given level of thymic output . In such models , following transient depletion , T-cell numbers will eventually rebound to those in healthy age-matched animals , irrespective of the extent of depletion ( Fig 6A ) or the age of the animal ( Fig 6B ) . This recovery happens on a timescale dictated by the rate of turnover . Here , the naive pool is essentially ‘memoryless’ , in the sense that there is no imprint of the pool’s developmental history on its potential for recovery . In contrast , in a purely adaptive model of homeostasis , there is no compensatory proliferative renewal of existing cells under lymphopenia , and cells behave essentially independently and according to their post-thymic age . In this regime , there is no unique set point for naive T-cell numbers . The extent of recovery is dictated by the extent of depletion ( Fig 6C ) and is further limited in older mice due to waning thymic output ( Fig 6D ) . We would expect the latter effect to be even stronger in humans , who experience a more substantial drop in thymic output with age [8 , 34–36] . Indeed , the correlation between loss and recovery predicted by the adaptation model ( Fig 6C ) is strikingly similar to the observation that pretreatment numbers of naive CD4 T cells in HIV-infected patients are a strong predictor of the recovery of the CD4 T-cell pool following antiretroviral therapy [37–40] . The predictions in Fig 6D also echo the observation that the potential for restoration of naive T-cell numbers in HIV-infected patients is progressively impaired with age [38 , 41 , 42] . While compensatory proliferation does occur in severely T cell–depleted humans , we speculate that the history of the T-cell pool , reflected in its age structure , may also be a determinant of its capacity for reconstitution .
Discussions of naive T-cell dynamics often focus on the ability of these cells to sense ‘space’ and that competition between and within TCR clones for resources determines the pool size and shapes its diversity [7 , 43–46] . However , it is clear that there is also heterogeneity in the homeostatic fitness of individual T cells that may relate to their developmental status in the periphery , intrinsic variation programmed during ontogeny , or to changes induced by interactions with their microenvironment over time . Given this heterogeneity , we questioned here whether quorum-sensing is indeed the dominant regulatory mechanism under normal physiological conditions and asked whether any one of these homeostatic forces can best account for multiple readouts relating to the long-term dynamics of the naive pool . Remarkably , a model of continuous adaptation , in which homeostatic fitness increases with cell age , emerged as the only mechanism capable of explaining all of the datasets ( Table 2 ) . While its visual descriptions of the data were consistently good , the model of adaptation was not statistically favoured as the best model in all cases . Nevertheless , the universality of its performance across multiple independent datasets weighs strongly in the model’s favour , particularly because alternatives that are strong competitors in one setting clearly fail to explain observations in another . Studying the kinetics of donor fractions in busulfan chimeric mice that are thymectomised at different times post BMT may provide additional empirical evidence to help distinguish the various models tested in this study ( S2 Fig , panel A ) because it would allow one to track population age structures at different levels of lymphopenia . We simulated such an experiment using parameters estimated from the fits to busulfan chimera data and found that the adaptation , selection , and incumbent models make distinct predictions regarding the dynamics of chimerism within the naive pools ( S2 Fig , panel B ) . While the model of adaptation in isolation proved most robust when fitted to diverse datasets , we do not exclude it acting in combination with other homeostatic mechanisms . Selection for fitter cells alone fails to explain all the data , but given the natural variation in TCR affinity for self-peptide–MHC , we might expect some level of accumulation of cells at the upper end of acceptable self-reactivity with age . Furthermore , the data we analysed in this study related to the natural history of naive T-cell numbers across the full lifespan of healthy mice and the partially depleted compartments induced by thymectomy in young adulthood . Quorum-sensing did not need to be invoked over this wide range of cell numbers , but there is evidence for it in more profoundly lymphopenic conditions [11 , 13 , 47–51] , and so it likely acts in concert with adaptation under these conditions . T-cell receptor excision circles ( TRECs ) are nonreplicating circular fragments of DNA that result from TCR rearrangement in the thymus and are distributed at random to daughter cells during mitosis . Their frequency in a naive T-cell population is determined by their rate of influx from the thymus and the diluting effects of cell division . TRECs therefore are potentially informative regarding homeostatic mechanisms . den Braber et al . [8] found ( i ) no significant decline in TREC frequencies within splenic naive CD4 and CD8 T cells with mouse age and ( ii ) a drop in TREC content following thymectomy in naive CD8 T cells ( spleen ) and both naive CD4 and CD8 T cells ( lymph nodes ) . These observations can be reconciled if one takes into account the constant replenishment of TREC-rich naive T cells from the thymus in healthy mice , which exceeds the rate of cell production by peripheral division . In healthy mice , any reduction of TREC frequencies through division , or by the preferential accumulation of relatively TREC-depleted older cells through increased survival , will then be countered by the influx of RTEs , which remains considerable across a mouse lifetime . The first observation therefore gives us little information with which to distinguish competition and adaptation . Similarly , TREC decline in Tx mice can be explained by either mechanism . With adaptation , the combination of slow peripheral division but preferential survival of older , TREC-depleted , naive T cells drives down the average TREC content when the TREC-rich RTE supply is removed . On similar lines , Thomas-Vaslin et al . [49] studied an experimental system in which apoptosis was chemically induced in dividing cells both in the thymus and the periphery . In otherwise healthy mice , they found that the subsequent approximately 50% drop in naive T-cell numbers was equivalent to that incurred by thymectomy , but the same treatment in Tx mice had no impact on naive T-cell numbers over a 2-wk period . These observations are consistent with the consensus that the thymus is the dominant source of naive T-cell production in mice and that there is little or no increase in peripheral division to compensate for thymectomy [52] . Thomas-Vaslin et al . found that in healthy mice , naive T-cell numbers returned to normal levels within 10 wk of treatment . We found that with an active thymus , numbers can be restored to close to healthy levels in a model of adaptation within a similar time frame ( Fig 6C ) . The combination of these observations lends further weight to the explanatory power of the adaptation model under replete or partially lymphopenic conditions . Our models do not address the underlying mechanism driving adaptation , though it seems likely to be due to the accumulation of signals from the cells’ environment , similar to other biological systems in which recurrent signals lead to cellular adaptation . T cells might also modify their sensitivity to homeostatic cues in response to regular interactions with self-peptide–MHC ligands and cytokines such as IL-7 [53–55] . Whatever the causal mechanism , survival may be the key property that is subject to adaptation in mice . Progressive alterations in intracellular prosurvival factors such as Bim and Erk kinases have been implicated in increasing the longevity of naive T cells [14 , 56] , while levels of homeostatic proliferation in mice do not appear to increase with age [13] . Overall , then , we argue from parsimony that the main determinants of naive T-cell numbers in healthy mice are ( i ) thymic output , which is very significant in early life but declines with age; ( ii ) low and relatively constant levels of renewal by homeostatic proliferation; and ( iii ) a gradual increase in cell longevity with post-thymic age . This is an especially important conclusion in relation to comparisons of adaptation and selection; these are fundamentally distinct processes , but both result in an increase of average cell fitness with time and are difficult to distinguish directly with experiments alone . In humans , proliferation of naive CD4 and CD8 T cells , as measured directly by Ki67 levels , increases with age [36] . Increasing proliferation might be driven by reduced competition from RTEs , whose influx wanes rapidly from adulthood onwards . A nonexclusive possibility is that naive T cells’ intrinsic propensity for proliferation , rather than survival , may increase with cell age in humans . This mechanism has been invoked to explain the relatively sudden loss of naive TCR diversity in the elderly [15] . Whether increasing homeostatic fitness with cell age is beneficial in evolutionary terms is not known . Naive T cells undergo profound changes in their function with host age , showing diminished activation and proliferation in aged mice and humans [56–58] . This development of functional defects appears to be a cell-intrinsic process rather than an effect of the aged environment . Parking of naive CD4 T cells in WT mice for different durations showed that old , longer-lived cells proliferate poorly and produce lower amounts of IL-2 in response to cognate antigen [54] , and new naive T cells generated in old mice using bone marrow chimeras exhibit normal function [59 , 60] . Despite this loss of function with cell age , it is possible that progressively increasing naive T cells’ ability to persist in the pool helps maintain a sizeable and diverse T-cell repertoire as thymic output wanes . Tuning this persistence via increased survival rather than division is also perhaps desirable because it seems more likely to achieve both stability of cell numbers and maintainance of TCR diversity as replenishment declines . Adaptation as passive accumulation of cells may therefore be an optimal means of making the best of our immune systems as we age . A potential pitfall arises , however , if resource competition plays any additional role in normal naive T-cell homeostasis . In this case , aged and impaired naive cells may actively outcompete younger , more functional ones—an effect that would contribute to the decline in immune responsiveness in the elderly . | The body maintains large populations of naive T cells , a type of white blood cell that is able to respond specifically to pathogens . This arsenal is essential for our capacity to fight novel infections throughout our lifespan , and their numbers remain quite stable despite a gradual decline in the production of new naive T cells as we age . However , the mechanisms that underlie this stability are not well understood . In this study , we address this problem by testing a variety of potential mechanisms , each framed as a mathematical model , against multiple datasets obtained from experiments performed in mice . Our analysis supports a mechanism by which naïve T cells gradually increase their ability to survive the longer they reside in the circulation . Paradoxically , however , naïve T cells may also lose their ability to respond effectively to infections as they age . Together , these processes may drive the accumulation of older , functionally impaired T cells , potentially at the expense of younger and more immunologically potent cells , as we age . | [
"Abstract",
"Introduction",
"Materials",
"and",
"methods",
"Results",
"Discussion"
] | [
"blood",
"cells",
"medicine",
"and",
"health",
"sciences",
"thymectomy",
"immune",
"cells",
"cell",
"cycle",
"and",
"cell",
"division",
"immunology",
"cell",
"processes",
"homeostatic",
"mechanisms",
"surgical",
"and",
"invasive",
"medical",
"procedures",
"physiological",
"processes",
"homeostasis",
"cytotoxic",
"t",
"cells",
"endocrine",
"system",
"procedures",
"white",
"blood",
"cells",
"animal",
"cells",
"t",
"cells",
"thymus",
"immune",
"system",
"cell",
"biology",
"natural",
"selection",
"physiology",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"evolutionary",
"biology",
"evolutionary",
"processes"
] | 2018 | Age is not just a number: Naive T cells increase their ability to persist in the circulation over time |
During C . elegans development , microRNAs ( miRNAs ) function as molecular switches that define temporal gene expression and cell lineage patterns in a dosage-dependent manner . It is critical , therefore , that the expression of miRNAs be tightly regulated so that target mRNA expression is properly controlled . The molecular mechanisms that function to optimize or control miRNA levels during development are unknown . Here we find that mutations in lin-42 , the C . elegans homolog of the circadian-related period gene , suppress multiple dosage-dependent miRNA phenotypes including those involved in developmental timing and neuronal cell fate determination . Analysis of mature miRNA levels in lin-42 mutants indicates that lin-42 functions to attenuate miRNA expression . Through the analysis of transcriptional reporters , we show that the upstream cis-acting regulatory regions of several miRNA genes are sufficient to promote highly dynamic transcription that is coupled to the molting cycles of post-embryonic development . Immunoprecipitation of LIN-42 complexes indicates that LIN-42 binds the putative cis-regulatory regions of both non-coding and protein-coding genes and likely plays a role in regulating their transcription . Consistent with this hypothesis , analysis of miRNA transcriptional reporters in lin-42 mutants indicates that lin-42 regulates miRNA transcription . Surprisingly , strong loss-of-function mutations in lin-42 do not abolish the oscillatory expression patterns of lin-4 and let-7 transcription but lead to increased expression of these genes . We propose that lin-42 functions to negatively regulate the transcriptional output of multiple miRNAs and mRNAs and therefore coordinates the expression levels of genes that dictate temporal cell fate with other regulatory programs that promote rhythmic gene expression .
MicroRNAs ( miRNAs ) are non-coding RNA molecules that post-transcriptionally regulate gene expression [1] . The maturation of miRNAs is a stepwise process that begins with the RNA polymerase II-dependent transcription of long capped and polyadenylated primary miRNAs ( pri-miRNAs ) [2] , [3] . Most pri-miRNAs are then endonucleolytically cleaved by the nuclear Microprocessor complex , composed of Drosha ( an RNase III enzyme ) and its binding partner Pasha , to yield a ∼70 nt precursor miRNA hairpin ( pre-miRNA ) [4] . After export to the cytoplasm , the pre-miRNA is cleaved by Dicer ( a second Type III RNase ) yielding a ∼22 nt duplex that consists of the mature miRNA and its corresponding passenger RNA [5] , [6] . The mature single-stranded ∼22 nt miRNA is then loaded into the Argonaute and GW182 to form the miRNA-induced Silencing Complex ( miRISC ) [7]–[9] . Through partial complementary base-pairing between the miRNA and target mRNA , the miRISC complex negatively regulates gene expression by either translational repression or mRNA degradation [7] , [10] . In vivo , target mRNA down-regulation is directly proportional to the amount of miRNA associated with miRISC [1] . Experimental and computational approaches indicate that an individual miRNA can bind to and regulate hundreds of mRNAs and that the majority of protein-coding genes are miRNA targets [11]–[14] . As such , miRNAs have been implicated in a variety of developmental and cellular processes including cell fate specification , proliferation and apoptosis [15]–[19] . In many of these scenarios , the expression of distinct miRNAs is tightly controlled and/or the individual steps of miRNA biogenesis are actively regulated at either the transcriptional or post-transcriptional level by sequence-specific transcription factors or RNA-binding proteins , respectively . For example , some regulatory proteins control miRNA biogenesis by directly binding structural elements within the pri- or pre-miRNA transcript whereas others broadly impact global miRNA biogenesis by inhibiting enzymes required for general miRNA processing and/or activity [20] . Importantly , many of the proteins that regulate miRNA biogenesis are highly conserved and mutations in these genes result in a variety of developmental disorders and diseases [20] . The C . elegans heterochronic pathway has been instrumental to our understanding of the principles of miRNA-mediated gene regulation and for the identification of components that are required to control miRNA expression , metabolism and activity [21] . Post-embryonic development in C . elegans proceeds through a series of four larval stages , punctuated by molts , in which the temporal and spatial patterns of cell division and differentiation are tightly orchestrated and invariant [22] . Heterochronic genes organize temporal patterns of development by controlling stage-specific gene expression . Defects in heterochronic genes cause animals to display temporal cell fate transformations including either the inappropriate skipping or reiteration of stage-specific patterns of cell divisions [23] . An overarching feature of the heterochronic pathway is that many protein-coding genes that are important for controlling temporal patterning are post-transcriptionally regulated by miRNAs [16] , [24]–[28] . In this context , miRNAs are expressed at defined times during post-embryonic development and function as molecular switches to inhibit earlier patterns of development and promote the emergence of later gene expression profiles . Throughout post-embryonic development , the expression of heterochronic miRNAs is regulated at both the transcriptional and post-transcriptional levels [20] , [29]–[32] . In addition , mutations that alter heterochronic miRNA expression often display strong temporal patterning and behavioral phenotypes [16] , [33]–[36] . While the regulatory strategies that dictate patterns of cell fate specification have rapidly emerged through the identification of conserved heterochronic genes , we still lack a deep understanding of how the temporal expression of heterochronic genes are coordinated with aspects of growth and behavior . This coupling is especially important as many post-embryonic cell division and cell fate specification events are intimately tied to the molting cycle [37] , [38] . Surprisingly , most of the known genes required for molting do not dramatically alter temporal cell fates and only a few heterochronic genes disrupt the reiterative process of molting [23] , [38]–[45] . The molting phenotypes associated with heterochronic mutants usually result from inappropriate temporal cell fate transformations that lead to either a cessation ( for precocious heterochronic mutants ) or an inappropriate reiteration ( for retarded heterochronic mutants ) of molting [16] , [23]–[28] , [31] , [33] , [34] , [42]–[45] . To date , only a single heterochronic gene , lin-42 , is known to alter both temporal patterning of cell fate specification and the precise timing of recurrent developmental events [39] . lin-42 is the C . elegans homolog of human and Drosophila PERIOD and was initially identified as a heterochronic mutant that precociously executes adult-specific patterns of development after the third larval molt [46]–[48] . The lin-42 locus is complex and encodes three protein isoforms ( LIN-42A , LIN-42B and LIN-42C ) that are expressed from two distinct promoters ( Figure 1A ) [39] , [46]–[48] . During post-embryonic development , lin-42 mRNA levels fluctuate over the molting cycles and peak once during each larval stage [39] , [46]–[48] . While its precocious developmental phenotypes are similar to other heterochronic mutants , the periodic expression pattern of LIN-42 distinguishes it from other monotonically expressed heterochronic proteins . Therefore , lin-42 has been proposed to play a more iterative role in developmental timing . However , its relationship to and interplay with other heterochronic genes has been difficult to establish at the molecular level . In addition to altering temporal patterns of development , mutations that disrupt the expression of LIN-42A and LIN-42B isoforms display dramatic defects in behavior and molting [39] . Specifically , lin-42a/b mutants alter the normally synchronous molting patterns displayed by wild-type animals and these defects frequently result in lethality [39] . Given that LIN-42 is a nuclear protein , an attractive hypothesis is that LIN-42 coordinates gene expression programs that control the molting cycles with regulatory pathways that mediate stage-specific cell lineage programs [48] . However , this potential role for LIN-42 remains elusive because 1 ) the molecular nature of LIN-42 activity is yet to be defined and 2 ) LIN-42 downstream targets that mediate iterative ( molting ) and sequential ( cell fate patterning ) gene regulatory programs are unknown . In this study , we employed multiple forward genetic screens that were collectively geared to identify negative regulators of miRNA expression . As a product of this approach , we identified mutations in lin-42 that suppress multiple stage-specific lineage defects associated with heterochronic miRNAs . Analysis of miRNA expression in lin-42 mutant animals suggests that LIN-42 broadly functions to negatively regulate miRNA expression is therefore is likely to act in a variety of pathways that require miRNAs for proper cell fate specification . Consistent with this hypothesis , we find that lin-42 also plays a role in the miRNA-mediated specification of asymmetric gene expression patterns in gustatory neurons . Analysis of LIN-42 interactions with chromatin suggests that LIN-42 potentially regulates the transcription of both miRNAs and mRNAs . We demonstrate , through the use of transcriptional reporters , that lin-42 mutations alter the transcription of lin-4 and let-7 . Surprisingly , mutations that remove LIN-42 isoforms containing the conserved PAS domains ( required for circadian gene regulation by human and Drosophila PERIOD ) do not uncouple miRNA expression from the molting cycle but , instead , dramatically alter the transcriptional output of miRNA genes . We conclude that a key molecular function of lin-42 is to dynamically inhibit the transcription of post-embryonically expressed miRNAs and mRNAs to ensure the robustness of developmental gene expression .
The inherent dependency of the heterochronic pathway on precisely controlled miRNA activity provides a unique genetic context to identify components that control aspects of miRNA metabolism or expression . To accomplish this , we performed forward genetic screens in either lin-4 ( ma161 ) , alg-1 ( ma192 ) or let-7 ( n2853 ) mutant backgrounds to identify novel heterochronic mutations that correct the phenotypes associated with aberrant L1 to L2 ( early ) , L2 to L3 ( middle ) or L4 to adult ( late ) cell fate transitions , respectively . These mutants are unique in that they express miRNAs at a much lower level than wild-type animals but do not completely eliminate their expression . lin-4 ( ma161 ) and let-7 ( n2853 ) mutations alter the conserved seed sequence of the mature miRNA and reduce levels of these miRNAs in vivo [16] , [24] . Animals harboring lin-4 ( ma161 ) and let-7 ( n2853 ) mutations are phenotypically indistinguishable from null mutants and reiterate L1- and L4-specific cell fates , respectively ( Table 1 ) [16] , [24] . alg-1 ( ma192 ) mutations alter one of the two miRNA-specific Argonautes and disrupt the ability of processed miRNAs to repress downstream target mRNAs [49] . Animals harboring the alg-1 ( ma192 ) mutation inappropriately express hbl-1 ( the major miRNA target of miR-48 , miR-241 , and miR-84 ) in the L3 stage and reiterate L2-specific seam cell division patterns [49] . Consistent with the defects associated with the misregulation of each of these stage-specific transitions , lin-4 ( ma161 ) , alg-1 ( ma192 ) and let-7 ( n2853 ) animals display highly penetrant heterochronic phenotypes and fail to express adult-specific gene regulatory programs , including the expression of the adult-specific Pcol-19::GFP transcriptional reporter ( Table 1 ) . Suppressors of the retarded heterochronic phenotypes in each of these genetic backgrounds were identified as F2 progeny of mutagenized animals that were able to restore normal development ( Figure 1C ) . Five mutants ( ma206 , ma208 , csh1 , csh4 and csh5 ) were able to suppress multiple retarded heterochronic phenotypes associated with all three mutant backgrounds . Each mutant mapped to a single locus on chromosome II ( Figure 1A and Table 1 ) , and subsequent SNP-SNP mapping and sequencing results demonstrated that all five alleles contain mutations that lie within lin-42 and would be predicted to create a premature truncation of the lin-42b or lin-42c open reading frames ( Figure 1A and Table S1 ) [50] . Consistent with previous analyses of lin-42 mutations , animals harboring the ma206 , ma208 , csh1 , csh4 or csh5 allele display highly-penetrant precocious heterochronic phenotypes ( Table 1 ) which were rescued with a fosmid containing the genomic fragment of the wild-type lin-42 gene [46]–[48] . Mutations in lin-42 have been demonstrated to suppress heterochronic phenotypes associated with multiple heterochronic mutants , including lin-4 and let-7 [16] , [46] , [47] , [51] . In these reports , only terminal cell lineage phenotypes , including a correction of the L4-to-adult vulval bursting phenotypes , restoration of adult-specific expression of Pcol-19::GFP , and formation of adult-specific alae were assayed . We next sought to determine whether the lin-42 mutations we isolated suppressed only terminal heterochronic phenotypes or if they corrected additional stage-specific cell lineage defects associated with lin-4 ( ma161 ) , let-7 ( n2853 ) and alg-1 ( ma192 ) mutations . To test if our new lin-42 mutants correct retarded cell lineage phenotypes , we compared multiple hypodermal cell lineages in lin-4 ( ma161 ) , alg-1 ( ma192 ) , and let-7 ( n2853 ) single mutants to double mutants that also harbored the individual lin-42 candidate suppressor mutations . lin-4 animals lack vulval structures as a consequence of reiterating L1-specific developmental programs in the hypodermis and failing to interpret inductive cues from the anchor cell that initiate vulval morphogenesis at the L3 stage [40] , [52] . The vulvaless ( Vul ) phenotypes of lin-4 ( ma161 ) animals are highly penetrant ( Figure 1B , F ) and are almost completely suppressed by lin-42 ( ma206 ) , lin-42 ( ma208 ) , lin-42 ( csh1 ) , lin-42 ( csh4 ) and lin-42 ( csh5 ) ( Figure 1B , F ) . These results indicate that lin-42 functions to control cell fate specification in at least the mid-L3 stage , when the vulval precursors are spatially patterned . The ability of several of these suppressors to alleviate hypodermal cell lineage phenotypes in miRNA hypomorphic mutants was not limited to the vulval cell lineage . The lateral seam cells of lin-4 ( ma161 ) , alg-1 ( ma192 ) , and let-7 ( n2853 ) animals display altered temporal cell fate specification and also fail to terminally differentiate at the L4 molt . As a consequence , lin-4 ( ma161 ) , alg-1 ( ma192 ) , and let-7 ( n2853 ) animals lack alae structures as young adults ( Table 1 , Figure 1D ) . The alae phenotypes in lin-4 ( ma161 ) , alg-1 ( ma192 ) , and let-7 ( n2853 ) mutants was strongly suppressed by the lin-42 ( ma206 ) allele ( Table 1 ) . alg-1 ( ma192 ) mutants reiterate L2-specific seam cell division programs due to the inappropriate perdurance of hbl-1 expression at the L3 stage [49] . As a consequence , young adult alg-1 ( m192 ) animals harbor supernumerary seam cells ( 23 . 5+/−3 . 78; WT = 11 ) ( Figure 1D , E ) . lin-42 ( ma206 ) mutations strongly suppress the L2-to-L3 heterochronic phenotypes of alg-1 ( ma192 ) mutants as lin-42 ( ma206 ) ; alg-1 ( ma192 ) animals exhibit a significant reduction in the number of supernumerary seam cells ( 11 . 9+/−1 . 3 ) and display normal adult alae ( Figure 1D and E ) . Therefore , lin-42 has a role in controlling L2-to-L3 temporal cell fate transitions . We asked whether our new lin-42 alleles could suppress the heterochronic phenotypes associated with lin-4 ( e912 ) and let-7 ( mn112 ) null mutants to a level similar to that observed with the hypomorphic alleles used in our initial screens . To test this , we compared aspects of vulval cell proliferation and morphogenesis at the early L4 stage in lin-4 ( ma161 ) , lin-42 ( csh5 ) lin-4 ( ma161 ) , lin-4 ( e912 ) and lin-42 ( csh5 ) lin-4 ( e912 ) mutants to those of similarly staged wild-type and lin-42 ( ma205 ) animals . Lowering lin-42 function in the context of the hypomorphic lin-4 ( ma161 ) background results in a strong restoration of vulval development with 85% of animals exhibiting induction/proliferation and invagination of P cells from the larval cuticle ( Figure 1B and F ) . Surprisingly , 42% percent of lin-42 ( csh5 ) lin-4 ( ma161 ) animals exhibited morphologically normal adult vulva and were competent for egg laying ( n = 100 ) . In contrast , reducing lin-42 activity in lin-4 ( e912 ) animals has little or no effect on P cell proliferation and vulval morphogenesis ( Figure 1F ) . lin-42 exhibits a similar genetic relationship to let-7 mutations . Both hypomorphic ( n2853 ) and null ( mn112 ) alleles of let-7 display highly penetrant vulval bursting phenotypes at the L4-to-adult transition ( Figure 1G ) [16] , [28] . lin-42 mutations almost completely suppress the lethality associated with larval-to-adult transitions in let-7 ( n2853 ) animals but do not statistically improve the viability of let-7 ( mn112 ) adults ( Figure 1G ) . These results strongly suggest that lin-42 mutations are not bypass suppressors of lin-4 or let-7 mutant phenotypes but likely require a minimum level of lin-4 or let-7 activity for suppression . One mechanism by which lin-42 mutations could suppress multiple hypomorphic miRNA mutants would be that lin-42 normally functions to repress some aspect of miRNA metabolism . To directly test this hypothesis , we measured the abundance of several mature miRNAs when lin-42 function is compromised . Northern blot analysis of total RNA extracted from morphologically-staged , young adult animals demonstrates that the total amount of lin-4 and let-7 miRNAs in alg-1 ( ma192 ) mutants is 1–1 . 5 fold lower than the levels found in wild-type animals ( Figure 2A and B ) . In addition to reducing the levels of mature let-7 miRNA , alg-1 ( ma192 ) animals display a slight reduction in pre-miRNA processing and accumulate the pre-let-7 hairpin precursor . This under-accumulation phenotype of mature lin-4 and let-7 miRNAs in alg-1 ( ma192 ) mutants is suppressed when lin-42 function is compromised ( Figure 2A ) . Consistent with our hypothesis that lin-42 normally inhibits miRNA biogenesis , similarly-staged lin-42 ( ma206 ) mutants over-accumulate both lin-4 and let-7 miRNAs ( Figure 2A ) . While the amount of mature let-7 miRNA increases in lin-42 ( ma206 ) ; alg-1 ( ma192 ) double mutants , the ratio of pre-let-7 to mature let-7 miRNA is similar to that detected in alg-1 ( ma192 ) single mutants ( Figure 2A ) . Therefore , although mature let-7 miRNA over-accumulates in lin-42 ( ma206 ) mutants , there is no change in pre-let-7 to mature let-7 processing efficiency as compared to wild type . These data suggest that lin-42 mutations alter aspects of miRNA expression upstream of pre-miRNA processing . To determine if lin-42 plays a more broad role in modulating miRNA expression , we employed real-time quantitative PCR to measure the expression levels of additional miRNAs in morphologically-staged , young adult wild-type , lin-42 ( ma206 ) , alg-1 ( ma192 ) and lin-42 ( ma206 ) ; alg-1 ( ma192 ) animals . We measured a variety of miRNAs that display tissue-specific and temporal expression patterns that are distinct from lin-4 and let-7 miRNAs [35] , [53]–[58] . For comparison , we also assayed the expression of two additional small nuclear RNAs ( U18 and sn2343 ) as well as two 21U RNAs that associate with PRG-1 , a distinct Argonaute involved in the C . elegans piRNA pathway [59]–[61] . Consistent with the observation that alg-1 ( ma192 ) mutations broadly affect miRNA expression , the abundance of all miRNAs tested ( lin-4 , miR-48 , miR-241 , miR-84 , let-7 , miR-1 , miR-46 , miR-58 and miR-79 ) was decreased in alg-1 ( ma192 ) mutants ( Figure 2B ) . The general miRNA under-accumulation phenotype displayed in alg-1 ( ma192 ) mutants was suppressed by removing lin-42 function ( Figure 2B ) . Importantly , the expression levels of the 21U-RNA transcripts were not significantly altered in lin-42 ( ma206 ) mutants ( Figure 2B ) . Examination of miRNA expression in lin-42 ( ma206 ) mutants indicate that all tested miRNAs were overexpressed from ∼1 . 8 to ∼3 . 2 fold when compared to similarly-staged wild-type animals ( Figure 2B ) . miRNA stability is dependent on a variety of factors , including the expression levels of the Argonaute components of miRISC [62] . To determine if the increase in miRNA levels in lin-42 mutant backgrounds was due to the overexpression of the C . elegans miRNA-specific Argonautes ( ALG-1 and ALG-2 ) , we quantified the levels of functional ALG-1 and ALG-2 fluorescent reporters in animals with reduced lin-42 activity . The results of this analysis , presented in Figure S1 , indicate that ALG-1 and ALG-2 expression is not altered in lin-42 ( RNAi ) animals . Collectively , these results indicate that lin-42 functions to negatively regulate the expression of a wide range of miRNAs . Because lin-42 regulates the abundance of many miRNAs , we asked if lin-42 functions in other gene regulatory pathways where controlling the expression levels of specific miRNAs is critical for proper cell fate determination . To test this idea , we examined how mutations in lin-42 affected the cell fate specification of two bilaterally symmetric gustatory neurons , ASE left ( ASEL ) and ASE right ( ASER ) . Normally , a complex gene regulatory network composed of miRNAs and transcription factors form a bi-stable , double-negative feedback loop that ensures mutually exclusive gene expression programs in ASEL and ASER neurons [63] , [64] . A major determinant of the exclusive gene expression programs in these two neurons is the ASEL-specific expression of the lsy-6 miRNA and the resulting down-regulation of its target , cog-1 . Animals completely lacking lsy-6 fail to down regulate COG-1 in ASEL , and , as a consequence , ASEL neurons in lsy-6 ( ot71 ) null mutants adopt an ASER cell fate [63] . These phenotypes can be monitored by a failure to express the Plim-6::GFP transcriptional reporter in ASEL in lsy-6 mutants ( Figure 3A ) . Importantly , lsy-6-mediated repression of cog-1 is dosage-dependent; weak alleles of lsy-6 , such as ot150 , under-accumulate lsy-6 miRNA as a consequence of reduced lsy-6 transcription and result in a partially penetrant ASEL-to-ASER cell fate transformation phenotype ( Fig . 3B ) [64] . The ot150 allele of lsy-6 has been used in a variety of contexts as a sensitized genetic background to identify gene products that function in the miRNA pathway [65]–[67] . While 13% of animals harboring only the lsy-6 ( ot150 ) allele fail to maintain Plim-6::GFP in ASEL , the penetrance of this phenotype is partially suppressed in lin-42 ( ma206 ) ; lsy-6 ( ot150 ) double mutants ( Figure 3B ) , suggesting that lin-42 may play a modulatory role in neuronal cell fate specification . To further explore a potential role for lin-42 in assuring proper neuronal cell fate specification , we developed a more sensitive assay for the lsy-6-mediated repression of cog-1 . As previously mentioned , alg-1 ( ma192 ) mutants display defects in variety of miRNA-mediated processes , including developmental timing [49] . While the alg-1 ( ma192 ) mutation alone does not alter Plim-6::GFP expression in ASEL , combining alg-1 ( ma192 ) with lsy-6 ( ot150 ) results in a dramatic increase in ASEL to ASER cell fate mis-specification ( Figure 3B ) . As with the suppression of alg-1 ( ma192 ) heterochronic phenotypes , reducing lin-42 function significantly restores normal ASEL cell fate specification in lsy-6 ( ot150 ) ; alg-1 ( ma192 ) animals ( Figure 3B ) . Because lsy-6-mediated cell fate specification is established during embryonic development , we conclude that lin-42 functions throughout development and is critical for multiple miRNA-mediated developmental processes . To characterize the spatial and temporal expression patterns of lin-42-regulated miRNAs , we generated a series of engineered transcriptional reporters that contain between 2 and 5 kB of genomic upstream regulatory sequence that drives the expression of GFP fused to an optimized proline-glutamate-serine-threonine-rich ( PEST ) sequence . PEST domains have been demonstrated , in a variety of heterologous systems , to accelerate the degradation of target proteins via the nuclear and cytoplasmic 26S proteasome [41] , [68]–[71] . In contrast to transcriptional reporters that drive the expression of stable GFP , analysis of GFP-pest expression in Plin-4::GFP-pest , Plet-7::GFP-pest or PmiR-1::GFP-pest transgenic animals indicates that the expression of each transcriptional reporter is highly dynamic , with peak GFP-pest expression occurring once each larval stage ( n>30 animals per time point ) ( Figure 4A ) [29] , [53] , [55] . The highly dynamic nature of each expression pattern was then monitored in a population of worms that were transiently arrested at the L1 diapause and then developmentally synchronized by restoring bacterial food . For each of the mir::GFP-pest reporters , post-embryonic GFP-pest expression was first detected at approximately 14 hours ( Figure 4B , D and F ) . Once transcriptionally activated , Plin-4::GFP-pest and Plet-7::GFP-pest reporters peak in expression by 18–20 hours and diminish with similar kinetics ( Figure 4B and D ) . For animals expressing the Plet-7::GFP-pest reporter we monitored GFP-pest expression for longer periods after release from L1 arrest . Consistent with the highly pulsatile nature of this expression pattern , GFP-pest expression was reinitiated at 30 hours , which correlates with the later portions of the L2 stage ( Figure S3 ) . While transcriptional activation of the Pmir-1::GFP-pest reporter was also initiated at 14 hours post-L1 arrest , the peak of Pmir-1::GFP-pest expression occurred at a later time point , and diminished with slower kinetics , as compared to Plin-4::GFP-pest and Plet-7::GFP-pest expression ( Figure 4F ) . We then asked whether the temporal expression pattern of each Pmir::GFP-pest reporter was synchronized with defined stages of the molting cycle , specifically lethargus and ecdysis . To accomplish this , we isolated late-L3-staged transgenic animals and cultured them on separate nematode growth media ( NGM ) plates at 20°C . Individual animals were then monitored for GFP-pest expression in relation to the induction and termination of both lethargus and ecdysis ( Figure 4C , E , and G ) . We find that the majority of animals which harbor the Plin-4::GFP-pest transgene cease GFP-pest expression by L3 ecdysis and resume expression by the mid-L4 stage . The pulse of Plin-4::GFP-pest expression at the L4 stage extends through the early portion of young adulthood and completely overlaps with the lethargus period in all animals ( Figure 4C ) . Plet-7::GFP-pest expression followed a similar pattern ( Figure 4E ) . However , GFP-pest expression was more variable at the L3-to-L4 transition and L4-specific induction of this transgene was primarily restricted to the lethargus period ( Figure 4E ) . In contrast to the expression profiles of the lin-4 and let-7 reporters , induction of Pmir-1::GFP-pest expression began during , or immediately after , L3 ecdysis and persisted into the L4 stage . A second pulse of Pmir-1::GFP-pest expression completely overlapped with the L4 lethargus period and continued into early adulthood ( Figure 4G ) . Collectively , these results suggest that the expression patterns of lin-4 , let-7 and mir-1 are dynamic throughout development and that the cyclical transcription of these miRNAs is mediated by their cognate promoter sequences . Furthermore , these data show that , while each of the Pmir::GFP-pest reporters display pulsatile expression patterns , the transcriptional dynamics for each gene do not display a complete unity of phase in their expression profiles . To compare the temporal expression patterns of these three miRNAs with that of lin-42 , we constructed transgenic strains that expressed either Plin-42a::GFP-pest or Plin-42b::mCherry-pest and subjected these animals to the same time course analyses . It has been previously demonstrated that two independent promoters drive the expression of LIN-42A , LIN-42B and LIN-42C isoforms [39] . Consistent with these findings , Plin-42a::GFP-pest and Plin-42b::mCherry-pest reporters displayed highly pulsatile expression during the L1 stage with initiation and termination of expression at 12 and 28 hours post L1 arrest , respectively ( Figure 4H ) . In addition , we find that Plin-42a::GFP-pest expression peaks at 16 hrs , immediately preceding the expression of the Pmir::GFP-pest reporters , while the peak of Plin-42b::mCherry-pest expression occurs at 20 hrs ( Figure 4H ) . Detailed analysis of individual L3-to-adult animals indicates that Plin-42a::GFP-pest expression displays a temporal expression pattern that is highly similar to Plin-4::GFP-pest and Plet-7::GFP-pest expression ( Figure 4I ) . Specifically , in all three reporters , GFP-pest expression diminishes prior to L3 ecdysis , resumes prior to the L4 lethargus period , and terminates immediately after L4 ecdysis ( Figure 4I ) . In striking contrast to our mir and lin-42 transcriptional reporters , Pcol-12::mCherry-pest expression does not occur during the molting cycle , but rather is exclusively expressed after each ecdysis ( Figures 4J and 4K ) . Previous analysis of lin-4 and let-7 expression indicates that these miRNAs are expressed in a variety of tissues , including the hypodermis , intestine and muscle [29] , [35] , [53] , [57] . To determine if Plin-4::GFP-pest displays differential temporal expression patterns in a subset of these tissues , we conducted a detailed examination of GFP-pest expression from the early-L3 to the young adult stage . Twenty animals from each of eight morphologically-defined stages were imaged ( Figure 4L and Figure S4 ) and then qualitatively scored for GFP-pest expression in seam cells , hyp7 cells or lateral muscle cells ( Figure 4M ) . Expression of the Plin-4::GFP-pest reporter peaked in hyp7 and seam cells at the mid- and late-L3 stage and then again at the late-L4 stage . In addition , the majority of animals exhibited a cessation of hyp7 and seam cell Plin-4::GFP-pest expression immediately after L4 ecdysis ( Figure 4M ) . In contrast , Plin-4::GFP-pest expression in muscle cells displayed a different transcriptional profile . In the majority of animals , expression of GFP-pest in muscles peaked at L3 ecdysis , gradually diminished throughout the remainder of the L4 stage , and increased again at the young adult stage ( Figure 4M ) . These results suggest that , while lin-4 is dynamically expressed once each larval stage , its promoter activity may be differentially regulated in distinct tissues . Analysis of the Plet-7::GFP-pest and Plin-4::GFP-pest reporters in lin-42 loss-of-function ( lf ) animals demonstrated that mutants that alter either lin-42 b/c ( lin-42 ( n0189 ) ) or lin-42a/b ( lin-42 ( ok2385 ) ) isoforms display elevated Pmir-GFP-pest expression in late larval development ( Figure 5A , B and Figure S2 ) . These altered temporal expression patterns suggested that lin-42 may normally function to modulate aspects of miRNA transcription . To investigate the potential interactions between LIN-42 and transcriptional regulatory elements , we performed chromatin immunoprecipitation coupled to high throughput sequencing ( ChIP-seq ) using extracts prepared from animals harboring a functional , GFP-tagged allele of lin-42 ( Figure 5C ) . From two independent biological ChIP-seq replicates derived from separate L4-staged extracts , we obtained 413 high confidence peaks corresponding to chromosomal regions in which LIN-42 is enriched ( see Table S2 and Materials and Methods ) . In agreement with the hypothesis that LIN-42 regulates let-7 transcriptional activity , we find LIN-42 binding sites at conserved let-7 promoter regions that have been previously demonstrated to control let-7 expression ( Figure 5C ) [31] , [35] , [57] . Annotation of additional high confidence peaks revealed that 38% ( 158/413 ) of LIN-42 peaks fell within the promoters ( defined as 2 kb upstream of each gene ) of either coding or non-coding genes , 24% ( 99/413 ) fell within the introns of coding genes , 8% ( 34/413 ) fell within gene bodies and 29% ( 121/413 ) fell within other intergenic regions ( Figure 5D ) . Comparison between LIN-42 peak frequency and their distribution relative to the closest annotated transcription start site ( TSS ) revealed that LIN-42 has two major regions of enrichment: 1 ) directly at TSSs and 2 ) at approximately 750 bp upstream of a TSS ( Figure 5E ) . Of these high confidence peaks , 323 were also detected in LIN-42 ChIP-seq samples obtained using an antibody against endogenous LIN-42 , suggesting that this list forms a short , but high confidence , group of LIN-42 target genes . A list describing all high-confidence annotated LIN-42 peaks is provided in Table S2 . Using the Generic Gene Ontology Term Mapper , we found that numerous genes with high-confidence LIN-42 peaks can be categorized into groups that function in many diverse biological processes , including development , transport , small molecule metabolism , embryogenesis and growth ( Table S3 ) . Collectively , these results strongly suggest that LIN-42 plays a role ( either directly or indirectly ) in a broad range of biological processes and that it predominately interacts with the promoter regions of coding and non-coding genes to regulate their expression . The genetic and regulatory relationships between lin-42 and lin-4 or let-7 , as well as the overlapping temporal expression patterns of these three genes , suggest that lin-42 may play a role in modulating the dynamics of lin-4 and let-7 transcriptional activity . To directly test the idea that lin-42 regulates miRNA levels at the transcriptional level , we quantified the transcriptional profiles of Plin-4::GFP-pest and Plet-7::GFP-pest reporters in wild-type animals and lin-42 ( n1089 ) mutants . The n1089 allele of lin-42 deletes genomic sequences that eliminate the coding potential of the lin-42b and c isoforms ( Figure 1A ) and displays strong heterochronic phenotypes [39] , [46] , [47] . Importantly , these isoforms contain the domains , PAS-A and PAS-B , that most closely link LIN-42 to PERIOD , a protein involved in controlling the cyclical expression patterns of circadian-regulated genes [39] , [47] , [48] , [72] , [73] . We focused on quantifying the GFP intensities of 1 ) the hypodermal cells in L3-to-adult-staged Plin-4::GFP-pest animals and 2 ) the seam cells of similarly-staged Plet-7::GFP-pest animals . These tissues and stages were selected for analysis because the majority of well-characterized heterochronic phenotypes are detected in these tissues [16] , [24] , [39] , [42] , [46]–[48] . Expression levels for each transcriptional reporter were analyzed throughout eight defined and sequential stages that spanned from early L3 to young adult ( Figure 6A–C and Figure S4 ) . In agreement with our previous observations , expression of Plin-4::GFP-pest in hypodermal cells of wild-type animals is dynamic throughout development and displays two main peaks of GFP expression: one at the late-L3 stage and the other at the L4 molt ( Figure 6A , B ) . Similar results are also observed in the seam cells of wild-type animals expressing the Plet-7::GFP-pest reporter ( Figure 6B , C ) . One exception , however , is that that the first peak of Plet-7::GFP-pest expression occurs at the mid-L3 stage ( Figure 6B , C ) . Surprisingly , we find that the cyclical pattern of expression of these reporters is not affected in animals carrying the lin-42 ( n1089 ) mutation; both lin-42 ( n1089 ) and wild-type animals display nearly identical Plin-4::GFP-pest and Plet-7::GFP-pest temporal expression patterns ( Figure 6A , B and C ) . In contrast , the abundance of GFP-pest expression for each reporter is universally higher in lin-42 ( n1089 ) mutants as compared to similarly-staged wild-type animals ( Figure 6 A , B and C ) . In the case of the Plin-4::GFP-pest reporter , higher levels of GFP-pest intensity are observed in hypodermal cells throughout all developmental stages , with the greatest difference occurring between the late-L3 and L3-molt stages ( 3 . 1 and 4 . 3 fold respectively ) ( Figure 6B ) . Interestingly , although Plet-7::GFP-pest expression in lin-42 ( n1089 ) mutants is also greater in seam cells between the late-L3 and L3-molt stages ( 2 fold each ) , Plet-7::GFP-pest expression in lin-42 ( n1089 ) and wild-type animals is practically indistinguishable from wild-type during the mid-L4 to the young adult stages ( Figure 6C ) . Taken together , these results suggest that mutations that abolish the expression of PAS domain-containing LIN-42 isoforms do not alter the cyclical expression patterns of miRNA genes during development . Rather , these mutations alter the transcriptional output of miRNAs that display oscillatory expression patterns . As demonstrated in Figure 5D and Table S2 , LIN-42 binds the putative regulatory regions of multiple protein coding genes . This observation raises the possibility that LIN-42 may modulate the transcriptional output of other developmentally regulated genes , including those whose expression , like that of lin-4 and let-7 , is also linked to the molting cycle . To determine if lin-42 mutants alter the transcriptional output of other cyclically expressed mRNAs , we observed the expression of two transcriptional reporters for genes involved in the molting process , Pmlt-10::GFP-pest and Pcol-12::mCherry-pest . In wild-type animals , Pmlt-10::GFP-pest transcription begins at the end of each larval period when the new cuticle is being synthesized [39] , [41] . We monitored the expression of Pmlt-10::GFP-pest in F1 animals that had been exposed to control RNAi or two RNAi constructs that target all major isoforms of lin-42 and induce precocious expression of Pcol-19::GFP and adult alae [74] . As with the expression of Plin-4::GFP-pest and Plet-7::GFP-pest reporters , the Pmlt-10::GFP-pest reporter maintained its normal , oscillatory pattern of expression in lin-42 ( RNAi ) animals . Quantification of Pmlt-10::GFP-pest reporter expression at the late-L4 stage ( where Pmlt-10::GFP-pest normally peaks [39] , [41] ) indicates that lin-42 depletion does not alter the transcriptional output of the mlt-10 promoter ( Figure 6D ) . In addition , quantification of the Pcol-12::mCherry-pest reporter in young adult lin-42 ( n1089 ) animals also indicates that mutations in lin-42 do not alter the temporal expression patterns or levels of the col-12 promoter ( Figure 6E ) . Therefore , while lin-42 mutations alter the transcriptional output of the lin-4 and let-7 genes , lin-42 does not play an essential role in controlling the oscillatory expression patterns or transcriptional output of all genes whose expression is tied to the molting cycle .
Using an unbiased genetic approach , we sought to identify factors that modulate the expression of miRNAs that are critical for controlling temporal patterns of development throughout post-embryonic development . Our strategy was two-fold: 1 ) we sought to identify suppressors of heterochronic miRNA mutant phenotypes characterized by stage-specific alterations in temporal patterning and 2 ) we focused on identifying suppressors that preferentially alleviate phenotypes that result from a reduction in , rather than a complete loss of , miRNA expression . These efforts identified lin-42 , the C . elegans homolog of the circadian period gene , as a component that not only modulates heterochronic miRNA expression , but also regulates the expression of a wide range of broadly expressed , and functionally distinct , C . elegans miRNAs . Previous genetic analyses implicated lin-42 as a heterochronic gene that normally inhibits the precocious expression of adult characteristics [39] , [46]–[48] . The precise placement of lin-42 in the developmental timing pathway has been difficult to incorporate due to the observation that lin-42 mutations alter cell lineage programs that occur exclusively in late development , namely the transition from the L3 to the L4 stage [46]–[48] , [51] . In addition , epistasis experiments with other developmental timing mutants suggest that its interaction with other heterochronic genes is complex [46] , [47] , [51] , [75] , [76] . Furthermore , unlike other components that control discrete aspects of temporal patterning and display monotonic expression patterns , lin-42 expression is highly dynamic , suggesting a reiterative role for it in the heterochronic pathway . Results from our screens have identified five new alleles of lin-42 that suppress the adult-specific gene expression defects of hypomorphic alleles of heterochronic miRNAs . We also find that lin-42 corrects stage-specific cell fate specification defects present throughout larval and adult development in these miRNA mutants . These results indicate that lin-42 functions iteratively to control temporal cell fate specification by controlling the transcription of distinct miRNAs . In addition , we demonstrate that our newly-identified lin-42 ( lf ) mutants precociously express adult-specific programs and that these defects are suppressed by mutations in components of the miRNA machinery . Accordingly , these data suggest that lin-42 ( lf ) heterochronic phenotypes are due to an overexpression of specific miRNAs that control temporal patterning . Also , our results demonstrate that lin-42 mutations are not bypass suppressors of the heterochronic phenotypes displayed by lin-4 and let-7 null mutants , suggesting that lin-42 suppresses retarded heterochronic phenotypes by increasing the expression of heterochronic miRNAs or enhancing their effectiveness in regulating miRNA targets . Multiple lines of evidence described in this manuscript support the conclusion that LIN-42 regulates the transcription of a wide array of miRNAs . First , we investigated how our lin-42 suppressor alleles affected the overall levels of a subset of miRNAs involved in developmental timing . alg-1 ( ma192 ) animals display profound defects in temporal cell fate specification and also under-accumulate both lin-4 and let-7 miRNAs . Genetic and molecular experiments indicate that lin-42 suppresses alg-1 ( ma192 ) -dependent phenotypes by increasing the available amount of mature miRNAs . Second , lin-42 ( ma206 ) mutants over-accumulate multiple miRNAs , including those with no apparent role in developmental timing . Consistent with the hypothesis that lin-42 functions in additional gene regulatory pathways that require miRNA activity , we demonstrated that lin-42 ( lf ) mutants suppress phenotypes associated with the under-accumulation of a miRNA that is essential for proper neuronal cell fate specification . Because lsy-6-mediated regulation of cog-1 expression is dosage-dependent , we speculate that lin-42 ( lf ) mutations suppress neuronal cell fate specification defects by de-repressing lsy-6 transcription in ASEL neurons . In order to understand how lin-42 may modulate miRNA expression , we pursued two lines of inquiry . First , we constructed a series of reporters that allowed us to measure , in detail , the transcriptional dynamics of multiple miRNAs in developing animals . Using these reporters , we found that several heterochronic miRNAs , such as lin-4 and let-7 , exhibit highly dynamic expression patterns that are synchronized with the expression of genes required for each molting cycle . Importantly , the expression of the Plin-4::GFP-pest and Plet-7::GFP-pest reporters coincided with the transcriptional activation of lin-42 . Further analysis of the lin-4 and let-7 reporters in a lin-42 mutant background indicated that one function of lin-42 is to negatively regulate the transcriptional output of miRNA promoters . Therefore , LIN-42 functions in a manner similar to the human and Drosophila PERIOD proteins , which inhibit the transcription of circadian regulated genes [77] , [78] . Second , it has previously been shown that LIN-42 is a nuclear protein , which suggests that it may play a role in directly regulating the pulsatile expression patterns of its downstream targets [48] . In order to explore potential roles for LIN-42 in directly controlling aspects of miRNA transcription , we performed ChIP-seq experiments to determine if LIN-42 interacts with the putative regulatory regions thought to control the expression of miRNAs and mRNAs . These experiments demonstrated that LIN-42 interacts with the promoters of non-coding genes ( including let-7 ) as well as protein-coding genes , suggesting that lin-42 may regulate the temporal expression of broad class of genes . Given the role of human and Drosophila period in regulating circadian gene expression , we were surprised to find that animals harboring the lin-42 ( n1089 ) allele , which abolishes the expression of PAS-containing lin-42 isoforms , maintained lin-4 and let-7 periodic expression patterns in later larval development . The PAS domains of human and Drosophila PERIOD are absolutely required to maintain the oscillatory expression patterns of circadian-regulated genes [72] , [77] , [78] . In our experiments , peak expression of the lin-4 and let-7 transcriptional reporters occurred at roughly the same developmental stages in both wild-type and lin-42 ( n1089 ) animals . Interestingly , although the temporal expression patterns were similar , the levels of each reporter were elevated ( as high as four fold ) in lin-42 ( n1089 ) mutants as compared to wild-type animals ( Figure 7A and B ) . Notably , lin-42 ( n1089 ) mutations do not alter the expression of the lin-42a isoform , which has been implicated in controlling the periodicity of the molting cycle [39] . While the dissection of lin-42 function will require further study , these findings are consistent with the modular nature of LIN-42 activities and suggest a novel role for the PAS domains of LIN-42 in regulating the transcriptional output of periodically expressed genes . Based on our current observations , we propose a model in which each of the lin-42 isoform functions to sculpt the dynamic transcription of both miRNAs and mRNAs . In out model , cis-regulatory elements within the promoters of specific miRNAs and mRNAs would be sufficient to drive periodic transcription . Regulatory elements within these sequences would be bound by a sequence-specific transcription factor ( TF ) that would promote the periodic transcription of these genes near the end of each larval stage ( Figure 7C ) . Based on the role of PERIOD in other organisms and our data demonstrating that LIN-42 binds to the putative cis-regulatory elements of several miRNAs and mRNAs , we propose a model in which distinct isoforms of LIN-42 function to regulate the activity of the TF at multiple , genetically separable levels . Our evidence suggests that mutations that specifically disrupt isoforms containing the PAS domain ( LIN-42B and LIN-42 ) , fail to properly limit the transcriptional output of genes regulated by the temporal specific TF ( Figure 7C ) . As a consequence , although these mutants display essentially normal temporal patterns of miRNA transcription , the elevated levels of heterochronic miRNAs lead to precocious developmental phenotypes . Importantly , mutations that only alter PAS-domain containing isoforms of LIN-42 retain the expression of LIN-42A ( Figure 7B ) [39] . Our model would also predict that mutations that disrupt LIN-42 isoforms that contain the conserved SYQ/LT domains ( LIN-42A and LIN-42B ) would have complex phenotypes with regard to periodic transcription . Indeed , animals harboring the lin-42 ( ok2385 ) allele , which disrupts the expression of the LIN-42A isoform ( containing the SYQ/LT domains only ) and deletes portions of the LIN-42B isoform ( containing both the PAS and SYQ/LT domains ) , precociously execute stage-specific gene expression , fail to maintain periodic molting cycles and overexpress Pmir::GFP-pest transcriptional reporters [39] ( Figure S2 ) . We interpret the complex phenotypes of lin-42 ( ok2385 ) animals as a reduction of the two modular activities of LIN-42 domains . Specifically , a reduction of LIN-42 PAS domain expression alters transcriptional output and deletion of LIN-42 isoforms which contain the SYQ/LT domains results in defects in periodic transcription . Further studies will be needed to define a specific molecular role for LIN-42 isoforms in maintaining normal periodic transcription . Recent reports suggest that a significant portion of the C . elegans transcriptome is dynamically expressed [79]–[81] . The combined interpretation of these studies suggests that the post-embryonic expression of 5–20% of mRNAs is synchronized with the molting cycles . The conservation of this process implies that these temporal gene expression patterns confer fitness to an organism and raise a number of interesting questions regarding the nature of developmental gene regulation [81] . Because many genes whose oscillatory expression patterns are coupled to the molting cycle control cell fate decisions and cell metabolism in a dosage-dependent manner , it is interesting to speculate how their temporal expression patterns , and levels , are coordinated with their targets . We suggest that lin-42 plays a fundamental role in this process for a wide range of non-coding and protein-coding genes . Because many of the transcriptional targets of lin-42 include miRNAs , each of which may regulate a vast array of genes , the impact on the dynamic nature of the C . elegans transcriptome during development may be immense .
C . elegans strains were grown under standard conditions and mutagenized as previously described [82] . Positional cloning of each suppressor was performed using standard methods [50] . Transformation of animals and integration of extrachromosomal arrays were performed as previously described [83] . See Text S1 for details of transgenic animals used in this manuscript . Lineage analysis and scoring of adult alae phenotypes were performed by picking staged animals of the indicated genotypes and monitoring seam cells derived from the V lineage as previously described [22] . All images were taken with an Axio Scope . A1 microscope equipped with a monochrome camera ( Diagnostic Instruments Inc ) and SPOT imaging software ( SPOT Imaging Solutions ) . GFP images of the hypodermal and seam cells were used for further quantification of Pmir::GFP-pest intensity . The average GFP intensity per area ( arbitrary units ) was quantified using ImageJ64 . For each reporter , 20 individual animals were analyzed per developmental stage . For Plin-4::GFP-pest , 10 hypodermal cell nuclei per animal per stage , or a total of 200 nuclei per time point , were used to calculate the average GFP intensity . For Plet-7::GFP-pest , 5 seam cells per animal per stage , or a total of 100 cells per time point , were used to calculate the average GFP intensity . Total RNA was isolated from staged populations of worms , and northern blots were performed as previously described [27] . Multiplex microRNA TaqMan assays were performed according to the manufacturer's specifications ( Life Technologies ) and quantified using the ABI 7900HT Fast Real-Time PCR system ( Applied Biosystems ) . For each biological replicate ( 3 total ) , the means and standard deviations of the raw Ct values were calculated and the representative heatmap demonstrating the fold change signal was created using R packages ( www . r-project . org ) . For the characterization of behavioral and GFP/mCherry reporter expression , animals were prepared in one of two ways . For analysis of L1-stage expression , embryos were bleached and staged according to standard protocols and then plated on standard NGM media with OP50 [84] . At indicated times after the release from L1 synchronization , L1-staged animals were imaged with an Axio Scope . A1 microscope . For analysis of the molting cycle and GFP/mCherry-pest reporter expression , individual animals ( non-motile , non-pharyngeal pumping ) were picked to fresh NGM plates seeded with 20 µL of OP50 . Time courses were initiated for each animal after each animal ecdysed . To determine the active and lethargic periods of animals at each stage , the pumping rates of individual animals were observed for 30 s of every hour . GFP/mCherry-pest expression was then monitored using a Zeiss SteREO Discovery V12 microscope with appropriate filters . To prevent photo-bleaching , each animal was exposed to <3 s of UV light . See Text S1 for details . | MicroRNAs play pervasive roles in controlling gene expression throughout animal development . Given that individual microRNAs are predicted to regulate hundreds of mRNAs and that most mRNA transcripts are microRNA targets , it is essential that the expression levels of microRNAs be tightly regulated . With the goal of unveiling factors that regulate the expression of microRNAs that control developmental timing , we identified lin-42 , the C . elegans homolog of the human and Drosophila period gene implicated in circadian gene regulation , as a negative regulator of microRNA expression . By analyzing the transcriptional expression patterns of representative microRNAs , we found that the transcription of many microRNAs is normally highly dynamic and coupled aspects of post-embryonic growth and behavior . We suggest that lin-42 functions to modulate the transcriptional output of temporally-regulated microRNAs and mRNAs in order to maintain optimal expression of these genes throughout development . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"developmental",
"biology",
"model",
"organisms",
"cell",
"biology",
"genetics",
"biology",
"and",
"life",
"sciences",
"molecular",
"cell",
"biology",
"research",
"and",
"analysis",
"methods"
] | 2014 | LIN-42, the Caenorhabditis elegans PERIOD homolog, Negatively Regulates MicroRNA Transcription |
A comprehensive understanding of the molecular machinery important for nociception is essential to improving the treatment of pain . Here , we show that the BMP signaling pathway regulates nociception downstream of the E3 ubiquitin ligase highwire ( hiw ) . hiw loss of function in nociceptors caused antagonistic and pleiotropic phenotypes with simultaneous insensitivity to noxious heat but sensitized responses to optogenetic activation of nociceptors . Thus , hiw functions to both positively and negatively regulate nociceptors . We find that a sensory reception-independent sensitization pathway was associated with BMP signaling . BMP signaling in nociceptors was up-regulated in hiw mutants , and nociceptor-specific expression of hiw rescued all nociception phenotypes including the increased BMP signaling . Blocking the transcriptional output of the BMP pathway with dominant negative Mad suppressed nociceptive hypersensitivity that was induced by interfering with hiw . The up-regulated BMP signaling phenotype in hiw genetic mutants could not be suppressed by mutation in wallenda suggesting that hiw regulates BMP in nociceptors via a wallenda independent pathway . In a newly established Ca2+ imaging preparation , we observed that up-regulated BMP signaling caused a significantly enhanced Ca2+ signal in the axon terminals of nociceptors that were stimulated by noxious heat . This response likely accounts for the nociceptive hypersensitivity induced by elevated BMP signaling in nociceptors . Finally , we showed that 24-hour activation of BMP signaling in nociceptors was sufficient to sensitize nociceptive responses to optogenetically-triggered nociceptor activation without altering nociceptor morphology . Overall , this study demonstrates the previously unrevealed roles of the Hiw-BMP pathway in the regulation of nociception and provides the first direct evidence that up-regulated BMP signaling physiologically sensitizes responses of nociceptors and nociception behaviors .
In spite of its clear medical importance , the molecular mechanisms of pain signaling remain poorly understood . Pain pathways in large part depend on sensory input from specialized sensory neurons called nociceptors [1] . Since the activation of nociceptors leads to pain sensation and the sensitization of nociceptors is thought to be a major contributor of pain pathogenesis , understanding the molecular mechanisms controlling nociceptor function is essential for improving the treatment of pain [2] . Drosophila melanogaster is a powerful model system for neurogenetic studies of nociception . Larval Drosophila show stereotyped behavioral responses to potentially tissue-damaging stimuli , such as noxious heat or harsh mechanical stimulation [3] . The most unambiguous larval nociception behavior involves a corkscrew-like rolling around the long body axis ( termed nocifensive escape locomotion ( NEL ) or simply “rolling” ) . Since rolling is specifically triggered by noxious stimuli and is clearly separable from normal larval locomotion , the analysis of NEL provides a robust behavioral paradigm to study nociception . Class IV multidendritic ( md ) neurons are polymodal nociceptors that are necessary for thermal and mechanical nociception in larvae [4] . Optogenetic activation of the Class IV neurons is sufficient for triggering NEL [4 , 5] . Accumulating evidence in studies of fly nociception suggests that the molecular pathways of nociception are conserved between Drosophila and mammals [3 , 6–15] . To identify genes important for nociceptor function , we recently performed thermal nociception screens in which we targeted the RNAi knockdown of nociceptor-enriched genes in a nociceptor-specific manner [16] . In this screen , we found that two RNAi lines targeting highwire ( hiw ) caused driver dependent hypersensitivity in thermal nociception assays ( revealed as a rapid response to a threshold heat stimulus ) indicating a potential role for hiw as a negative regulator of nociceptor activity [16] . hiw is an evolutionally conserved gene encoding an E3 ubiquitin ligase , whose function has been implicated in various aspects of neuronal development , synaptic function , and neuronal degeneration [17] . However , in contrast , very little is known about the roles of hiw in sensory processing and in controlling behavior . Here , we present additional and more specific evidence that hiw plays an important role in the regulation of behavioral nociception and nociceptor sensitivity through the bone morphogenetic protein ( BMP ) pathway .
To further investigate the potential function of hiw in nociception that was suggested by our previous study , we tested mutants for a strong loss-of-function allele of hiw ( hiwND8 ) in thermal nociception assays [18] . Unexpectedly , we found that genetic mutants of hiw showed insensitivity to a noxious temperature probe of 42 or 46°C , which was , surprisingly , the opposite of the previously described hiw RNAi phenotype ( Fig 1A ) [16] . Similar thermal insensitivity was also seen with other hiw alleles ( S1 Fig ) . Although hiw is widely expressed in the nervous system [18] , nociceptor-specific restoration of hiw expression rescued this insensitivity ( Fig 1A ) , indicating that hiw function in nociceptors is sufficient for restoration of normal thermal nociception and the relevant site of action was in nociceptors . Intrigued by the clear phenotypic distinction between genetic mutants and RNAi animals , we further dissected the nociception phenotype of hiw mutants by employing an optogenetic strategy . Optical activation of larval nociceptors via the blue light-gated cation channel Channelrhodopsin-2 ( ChR2 ) is sufficient to induce larval NEL [4 , 5 , 19] . Since nociceptor activation by ChR2 circumvents receptor potential generation but still depends on the machinery essential for downstream signaling ( Fig 1B ) , this technique has been utilized to distinguish genes that are important for primary sensory function from those that function in downstream aspects of signaling , such as action potential generation/propagation and/or synaptic transmission [10 , 20] . Using low intensity blue light ( 3 . 8 klux ) , which elicits NEL in about 20–30% of control animals expressing ChR2::YFP in nociceptors ( Fig 1C ) , we found that the hiwND8 mutants had a significantly increased probability to show NEL , indicating that the mutant for this allele is hypersensitive in response to optogenetic activation of nociceptors ( Fig 1C ) even though it was insensitive in thermal nociception assays . Tissue specific rescue experiments again showed that nociceptor specific expression of hiw was sufficient to rescue this optogenetic hypersensitivity ( Fig 1C ) . Taken together , these findings suggested that hiw has multiple , but dissociable , effects in the regulation of nociceptors . On the one hand , hiw regulated a sensory reception-dependent function causing insensitivity , but it also regulated a function downstream of sensory reception that caused hypersensitivity . Thus , the hypersensitivity seen in our earlier RNAi experiments is likely reflective of effects on the latter process . To further examine hiw’s role , we tested the effects of expressing hiw△RING in nociceptors . The hiw△RING transcript encodes a mutated form of hiw lacking the RING domain that is responsible for E3 ligase activity [21 , 22] . This mutated protein has been proposed to function as a dominant-negative poison subunit in multimeric Hiw E3 ligase complexes . Similar to our original observations with hiw RNAi , expression of hiw△RING in nociceptors resulted in significant hypersensitivity in thermal nociception ( Figs 1D and S2 ) . This manipulation also caused hypersensitive optogenetic nociception responses ( Fig 1E ) . As hiw encodes a large protein with many functional domains , and phenotypes of hiw mutants are known to show varied sensitivity to gene dosage [21] , the observed similarity between hiw△RING overexpression and hiw RNAi is suggestive of dosage-dependent effects of hiw in nociceptors . For instance , the dominant negative approach may lead to an incomplete loss of function for hiw that is similar to the effects of RNAi . It has been very recently shown that the canonical BMP pathway in nociceptors is required for nociceptive sensitization after tissue damage in Drosophila [23] . Since the BMP signaling pathway has also been proposed to be a downstream pathway regulated by Hiw in motoneurons [24] , we tested whether the BMP signaling pathway is regulated downstream of Hiw in nociceptors . We first examined the level of phosphorylated Mad ( pMad ) in nociceptor nuclei by quantitative immunohistochemistry , which is an established method for evaluating the activation level of intracellular BMP signaling [25–31] . In nociceptor nuclei , hiw genetic mutants showed significantly elevated pMad levels ( 33% ) in comparison to wild-type , even when processed together in the same staining solution ( see also Materials and Methods ) ( Fig 2A , 2B and 2F ) . A similarly modest change in pMad accumulation in motor neuron nuclei is associated with effects on presynaptic function and morphology at the neuromuscular junction ( NMJ ) [32 , 33] . An increased accumulation of pMad in the nucleus and the cytoplasm was observed in nociceptors expressing hiw△RING ( Fig 2C and 2F ) . Expression of wild-type hiw in nociceptors of hiw mutant animals rescued the elevated pMad level ( Fig 2D and 2F ) . We also confirmed that our immunohistochemistry successfully detected the increase of nuclear pMad caused by expressing the constitutively active form of thick veins ( tkvQD ) , which activates the intracellular BMP signaling cascade independently of BMP ligands [34] ( Fig 2E and 2F ) . These data together suggest that BMP signaling is negatively regulated downstream of hiw in larval nociceptors . In the larval motoneurons , it is known that pMad signals can be locally detected at synaptic boutons as well as nuclei [26 , 35 , 36] . However , in our samples no detectable pMad signals were observed at synaptic terminals in larval nociceptors ( Fig 2G ) . Next , we tested whether up-regulated BMP signaling in nociceptors is responsible for the hypersensitive nociceptive responses caused by hiw loss-of-function . mad1 encodes a dominant-negative form of Mad with disrupted DNA-binding ability [37] . When mad1 was expressed together with hiw△RING in nociceptors , the hypersensitive phenotype that was normally induced by the expression of hiw△RING alone was not detected ( Fig 2H ) . Since neither expressing Mad1 together with hiw△RING nor expressing Mad1 alone in nociceptors induced insensitivity to noxious heat ( S3 Fig ) , these results indicate that hypersensitive nociception caused by weak hiw loss of function requires an intact BMP signaling pathway that normally operates through Mad . This result is consistent with the elevated pMad observed with hiw loss of function as playing a causal role in the hypersensitive phenotypes . The MAP kinase ( MAPKKK ) wallenda ( wnd ) is a well-characterized target substrate of Hiw ligase [17] . Hiw negatively regulates the protein level of Wnd , and the Hiw-Wnd interaction is crucial for normal synaptic growth , but not for normal synaptic function in NMJ [31 , 38–40] . In addition , hiw interacts with wnd in Class IV neurons in the regulation of dendritic and axonal morphology [41] . In larval motoneurons , it has been suggested that wnd is not involved in the regulation of BMP signaling [31] . To test whether wnd is involved in the control of BMP signaling downstream of hiw in nociceptors , we examined a genetic interaction between hiw and wnd in double mutants . A wnd mutation in hiw mutant background did not suppress the elevated nuclear pMad level in nociceptors that we observed in the hiw mutant ( Fig 3A–3D and 3F ) , nor did wnd single mutants show altered nuclear pMad accumulation relative to controls ( Fig 3E and 3F ) . Interestingly , significant up-regulation of nuclear pMad signal was observed in nociceptors overexpressing wnd , but not with a kinase-dead version of wnd ( S4 Fig ) . Taken together , these results suggest that elevated nuclear pMad in hiw mutant nociceptors does not depend on the activity of Wnd , although overexpression of wnd with GAL4/UAS can cause elevated BMP signaling in nociceptors . To gain insight into which regions of Hiw protein are involved in attenuating BMP signaling in nociceptors , we performed an expression study of a series of Hiw dominant negatives with various deletions , which has been established by Tian et al . [39] ( Fig 4A ) . Expressing HiwNT ( N-terminal half of Hiw ) caused a greater than 200% increase in nuclear pMad signals compared to controls ( Fig 4B , 4C and 4H ) . HiwCT ( C-terminal half of Hiw ) and Hiw△RCC1 resulted in 99% and 68% increases in nuclear pMad signals , respectively ( Fig 4D , 4E and 4H ) . HiwCT and Hiw△RCC1 also caused marked accumulation of pMad signals in the cytoplasm of nociceptors ( Fig 4D and 4E ) , which was also observed with Hiw△Ring expression ( Fig 2C ) . This cytoplasmic accumulation of pMad signals is unlikely due to technical variability of immunostaining since the control samples treated in the same staining solutions with HiwCT or Hiw△RCC1 never developed such accumulations and cells nearby the nociceptors showed the normal pMad signal . In contrast , Hiw△HindIII and HiwCT1000 ( C-terminal only region of Hiw ) did not cause any changes in nuclear pMad signals in nociceptors ( Fig 4C , 4F and 4H ) . Thus , the attenuation of BMP signaling in nociceptors through Hiw appears to depend on different regions of Hiw from those that have been proposed to be involved in the regulation of NMJ morphology ( Hiw△RCC1 , and Hiw△HindIII function as dominant-negative in NMJ morphology while HiwNT and HiwCT1000 do not [39] ) . Because both HiwNT and HiwCT , which are largely non-overlapping N-terminal and C-terminal halves of Hiw , caused increased nuclear pMad signals , multiple regions of the Hiw protein must be intact for normal suppression of BMP signaling in nociceptors . Although a previous study by Follansbee et al . suggests that the canonical BMP signaling pathway in larval nociceptors is a necessary component for nociceptive sensitization after tissue-damage , whether up-regulation of BMP signaling in nociceptors is sufficient to sensitize nociception has not been proven and potential mechanisms leading to sensitization are unknown . Because our data support the notion that the up-regulation of BMP signaling in nociceptors plays a key role in inducing sensitized nociception , we tested whether up-regulation of intracellular BMP signaling in nociceptors is sufficient to induce nociceptive hypersensitivity . In thermal nociception assays , animals expressing the constitutively active BMP receptor tkvQD in nociceptors did exhibit significant hypersensitivity ( Figs 5A and S2 ) , and tkvQD also caused hypersensitive responses in optogenetic nociception assays . The latter suggests that elevated BMP signaling in nociceptors was able to sensitize nociception through a mechanism that was downstream of sensory reception ( Fig 5B ) . Although the dendritic structure of nociceptors in tkvQD overexpressing animals was not significantly altered ( Fig 5C–5E ) , overexpression of tkvQD caused overextension and overexpansion of nociceptor axon termini ( Fig 5F–5H ) . Combined , these data demonstrate that elevated BMP signaling in nociceptors is sufficient to sensitize thermal and optogenetic nociception behaviors in addition to causing increases in axon terminal branching . Since nociceptor-specific up-regulation of BMP signaling sensitizes thermal and optogenetic nociception behaviors , we next explored whether the up-regulation of intracellular BMP signaling actually sensitizes physiological responses of nociceptors . To observe neuronal responses of larval nociceptors to a range of thermal stimuli , we developed a preparation for optical recording from axon terminals of the nociceptive neurons . We then observed these terminals while we locally applied a thermal ramp stimulus to the larval body wall ( Fig 6A ) . To monitor Ca2+ , the genetically encoded sensor GCaMP6m was expressed under the control of ppk-GAL4 [42] . In control animals we observed a steep increase of the GCaMP6m signal in nociceptors when the ramping temperature reached the 39–47°C temperature range ( Fig 6B , 6B’ and 6D ) . We found that nociceptors expressing tkvQD showed a significantly greater increase of GCaMP6m signals through 36–50°C in comparison to those in controls ( Fig 6C , 6C’ and 6D ) , while basal fluorescence levels of GCaMP6m ( F0 ) were comparable between the control and tkvQD-expressing nociceptors ( Fig 6E ) . These results suggest that the significantly greater increase of GCaMP6m signals observed in nociceptors expressing tkvQD is due to the greater level of Ca2+ influx triggered by the heat ramp stimulus , and not to unintended transcriptional upregulation of GCaMP6m . Thus , elevated BMP signaling in nociceptors results in exaggerated Ca2+ signals at the terminals of nociceptors in response to heat in the noxious range . This conclusion is consistent with the behavioral nociceptive sensitization induced by the same intracellular up-regulation of BMP signaling in nociceptors . Chronic up-regulation of BMP signaling in nociceptors caused sensitization of behavioral nociception responses of larvae and an increased Ca2+ response of nociceptors to noxious heat , but also expansion of nociceptor terminals . To further separate the physiological and morphological effects of BMP up-regulation in nociceptors , we up-regulated BMP signaling during a shorter 24-hour time-window in larval stage . Using the temperature sensitive repressor of GAL4 activity ( GAL80ts ) [43] , we activated expression of tkvQD in larval nociceptors by shifting ppk-GAL4 UAS-Chr2::YFP tub-GAL80ts animals to 30°C for 24 hours . We then tested these larvae for sensitized optogenetic nociception . The 24-hour induction of tkvQD induced hypersensitivity in the optogenetic nocifensive responses and also significantly increased nuclear pMad levels relative to controls ( Fig 7A and 7B ) . However , no detectable axonal overgrowth was induced by 24-hour tkvQD expression ( Fig 7C and 7D ) . Unfortunately , we were not able to investigate the effects of this manipulation on nociception responses with a 39°C thermal stimulus because the prolonged incubation at 30°C interfered with 39°C NEL behavior in both controls and experimental animals ( S5 Fig ) . This latter finding indicates that the sensitivity of thermal nociception in Drosophila is modulated by the ambient temperature . Collectively , these data demonstrate that 24-hour activation of BMP signaling in nociceptors is sufficient to sensitize larval nociceptive response in the absence of detectable changes to axonal morphology . Taken together with our Ca2+ imaging results , these data suggest a physiological role for BMP signaling in the regulation of nociceptor sensitivity .
The data we present in this study suggest that hiw has at least two distinct functions in the regulation of nociceptor sensitivity . We found that strong loss-of-function mutants of hiw showed insensitivity to noxious heat but hypersensitivity to optogenetic stimulation of nociceptors ( Fig 1A and 1C ) . Since expressing wild-type hiw in nociceptors of hiw mutants rescued both phenotypes , loss of hiw in nociceptors is responsible for these two ostensibly opposing phenotypes ( Fig 1A and 1C ) . We also found that nociceptor-specific expression of hiwRNAi or hiw△RING caused only hypersensitivity ( Fig 1D and 1E ) [16] , indicating that the process that governs hypersensitivity is separable from the cause of insensitivity . As insensitivity was epistatic to hypersensitivity in thermal nociception assays , we used optogenetics to show that hypersensitivity is actually present in hiw genetic mutants as well as in previously described RNAi animals . The use of optogenetic stimulation of neurons allowed us to bypass the endogenous sensory reception step ( s ) and to reveal this role . Our data suggest that hiw is a ) required for the negative regulation of a neural pathway that is downstream of sensory reception and b ) required to confer normal sensitivity to noxious heat via sensory reception pathways . As strong hiw loss of function causes reduced dendritic arbors [41] while hiw RNAi does not [16] , it is possible that the reduced dendrite phenotype accounts for the insensitivity of the strong hiw alleles . Consistent with this hypothesis , many manipulations that cause insensitive thermal nociception are associated with a reduction in the dendritic arbor [16] . The phenotypic difference between strong loss-of-function mutants and RNAi or Hiw dominant-negative animals suggests that insensitive and hypersensitive phenotypes observed in hiw mutants have different sensitivity to the dosage of hiw . This has also been seen in the larval motor neuron system where it has been demonstrated that two different phenotypes of hiw in larval NMJ ( overgrowth of synaptic boutons and diminished synaptic function ) are separable by their different sensitivity to the dosage of hiw [21] . Our data also suggest that hiw may regulate distinct molecular pathways in motor neurons and in nociceptors . In the larval NMJ , mutations of hiw or expression of hiw△RING cause a diminished evoked excitatory junction potential ( EJP ) amplitude due to decreased quantal content in synaptic vesicles [18 , 21 , 46] . However , this diminished evoked EJP amplitude phenotype is apparently opposite to the hypersensitive nociception phenotype observed in this study . Thus , the downstream targets and/or pathways of Hiw in nociceptors may be distinct from those in motor neurons . We identified the BMP signaling pathway as an important signaling pathway in nociceptors that is regulated downstream of hiw . In fly motor neurons , it has been proposed that BMP signaling is a direct target of Hiw ligase [24] . However , a later study reported that pMad up-regulation was not detected in motor neuron nuclei in hiw mutants [31] and controversy has arisen over this interaction . We found that nuclear pMad signals were up-regulated in hiw mutant nociceptors , and that this molecular phenotype was rescued by wild-type hiw expression ( Fig 2 ) . In addition , we also detected striking accumulation of pMad in both the nuclei and cytoplasm of nociceptors expressing Hiw dominant negative proteins ( Figs 2 and 4 ) . Finally , using UAS-mad1 , we showed that a Mad-dependent pathway is responsible for the hypersensitive thermal nociception caused by hiw△RING expression ( Fig 2H ) . Our data therefore support the idea that the nociceptor BMP signaling pathway is regulated downstream from hiw . Although we demonstrated that BMP signaling is downstream of hiw in nociceptors , we have yet to determine the precise mechanism for Hiw regulation of BMP signaling . Our genetic analysis suggests that BMP signaling in nociceptors is regulated independently from the wnd pathway ( Fig 3 ) . Wnd is the best characterized target substrate of Hiw in the regulation of NMJ morphology [31 , 38–41 , 47] . Our expression analysis using various hiw deletion series showed that the set of hiw deletion constructs that induced up-regulation of BMP signaling in nociceptors was not identical to the set that induced abnormal synaptic morphology in motoneurons [39] . This finding is somewhat consistent with the existence of a Wnd-independent mechanism in the regulation of BMP signaling in nociceptors , since the Hiw-Wnd pathway plays a pivotal role in regulating synaptic morphology in larval NMJ . Intriguingly , our expression study of the hiw deletion series showed that the expression of HiwNT caused a prominent accumulation of nuclear pMad , while the expression of HiwCT or Hiw△RCC1 caused accumulation of pMad signals in both the nuclei and cytoplasm in nociceptors ( Fig 4C–4E ) . These data raise the possibility that Hiw is involved in at least two different mechanisms which regulate pMad: one pathway affecting nuclear pMad and another for cytoplasmic pMad . Given that hiw is a large protein with many functional domains for interacting with multiple molecules , the notion that hiw is involved in multiple processes regulating various aspects of neuronal functions in both motor neurons and nociceptive sensory neurons is perhaps unsurprising . Further studies are necessary to reveal the mechanisms of Hiw-dependent regulation of BMP signaling in nociceptors . We have presented a new physiological preparation for investigating the calcium levels in nociceptor terminals with a physiologically relevant noxious heat stimulus . This allowed us to demonstrate that up-regulation of BMP signaling in nociceptors sensitizes the physiological responses of nociceptors in response to noxious heat in addition to its effects on behavior ( Figs 5 and 6 ) . We also demonstrated that 24-hour activation of intracellular BMP signaling in nociceptors is sufficient for the nociceptive sensitization ( Fig 7 ) . Although it has been previously reported that BMP signaling in nociceptors is required for nociceptive sensitization after tissue-injury in Drosophila [23] , the mechanisms of the regulation of nociception by BMP signaling was totally unknown . Our study provides the first evidence to implicate BMP signaling in regulating physiological processes in nociceptors that control its sensitivity to noxious stimuli . The BMP signaling pathway plays crucial roles in various developmental processes , such as embryonic patterning , skeletal development , and the development of neuronal circuits [48 , 49] . The roles of BMP signaling in the regulation of neuronal activity has also been extensively investigated in larval motor neurons , where BMP signaling plays crucial roles in the homeostatic regulation of synaptic morphology and transmission [50 , 51] . In larval NMJ , the expression of active-form Tkv increases evoked EJP amplitude which is a similar effect on neuronal output to that we observed in nociceptors in this study [52] . A similar effect of active-form Tkv on evoked synaptic currents has been also reported in aCC interneurons in larval CNS [53] . These previous studies and this study together indicate that BMP signaling may function as a positive regulator of neuronal outputs . However , the previous studies and our current study also highlight differences in the functions of BMP signaling in different neurons . First , interfering with BMP signaling with dominant negative Mad did not cause nociception insensitive phenotypes ( S3 Fig ) ( consistent with another study that found that nociceptor-specific knockdown of BMP signaling components did not affect basal thermal nociception [23] ) . In contrast , loss of BMP signaling components in motor neurons decreased evoked EJP amplitude [24 , 36 , 54] . Second , expression of activated-Tkv in nociceptors resulted in an expansion of axonal projections ( Fig 5F–5I ) , the same manipulation does not increase the size of NMJ , while it increases nuclear pMad level also in motor neurons [24] . Although a full understanding of the mechanisms through which BMP signaling regulates nociceptor sensitivity requires further investigation , these results indicate that BMP signaling may act , at least in part , differently in the nociceptors and motor neurons to regulate neuronal outputs and morphology . Hiw and BMP signaling pathway components are all evolutionally well-conserved . The role of hiw in the negative regulation of nociceptive signaling may be as well . A mammalian hiw orthologue Phr1/MYCBP2 has been previously implicated in a negative regulation of nociception processing . Specifically , it has been reported that Phr1/MYCBP2 is expressed in DRG neurons , and that intrathecal injection of an antisense oligonucleotide against Phr1/MYCBP2 causes hypersensitivity in formalin-induced nociceptive responses [55] . Furthermore , nociceptive and thermoceptive neuron-specific Phr1/MYCBP2 knock-out mice show prolonged formalin-triggered sensitization in thermal nociception , whereas no obvious phenotypes are observed for basal nociception in the knock-out animals [56] . Decreased internalization of the TRPV1 channel ( which is mediated through a p38 MAPK pathway ) has been implicated in this prolonged nociceptive sensitization in MYCBP2 knock-out mice [56] . In contrast , whether BMP signaling plays a role in regulating nociception in mammals is unknown . Similarly , the degree to which the role of Hiw and BMP signaling is conserved in the physiological regulation of mammalian nociceptors represents a fascinating topic for future investigation . Intriguingly , Hiw and BMP signaling have been implicated in nerve regeneration/degeneration processes after axonal injury in both Drosophila and mammals [17 , 57] . In flies , axonal injury leads to decrease of Hiw , which leads to the upregulation of Wnd that promotes axonal degeneration in motor neurons [47] . Phr1/MYCBP2 is also involved in promoting axonal degeneration after sciatic or optic nerve axotomy [58] . Smad1 is known to be activated and play an important role for axonal regeneration after peripheral axotomy of DRG neurons [59–62] . Because nerve injuries are thought to be one of key contributors for neuropathic pain conditions and peripheral axotomies are widely used to generate neuropathic pain models in mammals , it will be of particular interest in the future to determine whether the Hiw-BMP signaling pathway and up-regulation of intracellular BMP signaling in nociceptors play a role in the development of a neuropathic pain state in mammals .
Canton-S and w1118 were used as control strains as indicated . The other strains used in this study were as follows: ppk1 . 9-GAL4 [63] , UAS-mCD8::GFP [64] , UAS-ChR2::YFP line C [4] , hiwND8 [18] , hiwΔN , hiwΔC , UAS-hiw , UAS-hiwΔRing [21] , UAS-hiwNT , UAS-hiwCT , UAS-hiw△RCC , UAS-hiw△HindIII , UAS-hiwCT1000 [39] , wnd1 , wnd2 , UAS-wnd [31] , ppk1 . 9-GAL4; UAS>CD2 stop>mCD8::GFP hs-flp , UAS-tkvQD [34] , tub-GAL80ts [65] , ppk-CD4-tdGFP [66] and UAS-GCaMP6m [42] . UAS-mad1 [37] The thermal nociception assay was performed as described previously [3 , 6 , 10 , 16 , 67] . NEL latency was measured as initial contact of the thermal probe on the lateral side of the larval body wall to the completion of NEL ( a 360° roll ) . Stimulation was ceased at 11 seconds . A thermal probe heated to 46°C was used to examine the insensitive phenotype since it usually evokes NEL in less than 3 seconds [3 , 6 , 10 , 16 , 68] . A 39°C probe , which usually results in NEL in 9–10 seconds , was used to examine thermal hypersensitivity , as using a lower temperature probe is important to detecting the hypersensitive phenotype [16] . The optogenetic nociception assay was performed as described previously [5] with slight modifications . 3 . 8 klux was used to test for optogenetic hypersensitivity , but 76 klux blue light was used in the analysis of 24-hour tkvQD induction ( Fig 7 ) . Because male larvae showed a lower responsiveness to optogenetic nociceptor activation than females ( S6 Fig ) , male larvae were used to allow for more easily detectable hypersensitivity . Antibodies used in this study were as follows: rabbit anti-GFP ( Invitrogen , 1:1000 ) , mouse anti-GFP ( Invitrogen , 1:250 ) , mouse anti-rat CD2 ( AbD Serotec , 1:200 ) , rabbit anti-pMad ( gift from Ed Laufer , 1:1000 ) , goat anti-rabbit Alexa488 ( Invitrogen , 1:1000 ) , goat anti-rabbit Alexa568 ( Invitrogen , 1:1000 ) , goat anti-mouse Alexa488 ( Invitrogen , 1:1000 ) and goat anti-mouse Alexa568 ( Invitrogen , 1:1000 ) . Larvae were filleted , fixed in 4% paraformaldehyde for 30 minutes and then stained according to a standard protocol [69] . Wandering third instar larvae expressing mCD8::GFP in nociceptors were filleted and immunostained as described above . To minimize variation due to processing controls , experimental specimens were processed side-by-side within the same staining solutions . In order to avoid skewing results from potential biases of pMad staining among different segments , one dorsal Class IV mutidendritic neurons ( ddaC ) each from segments A4 , 5 and 6 ( three neurons in total ) was imaged in each sample ( Zeiss LSM 710 with a 100x/1 . 4 Plan-Apochromat oil immersion or Olympus FV1200 with a 100x/1 . 4 UPLSAPO oil immersion ) . Z-stack images were converted to maximum intensity projections . To quantify nuclear pMad signals , nociceptor nuclei were identified based on the absence of GFP signal , and a region of interest ( ROI ) outlining the nucleus was delineated . The average signal intensity of nuclear pMad staining in the ROI was then calculated . Background signal intensity was determined as the mean from ROIs ( identical size and shape of the nucleus from the image ) drawn in the four corners of each image . The calculated background signal intensity was then subtracted from the nuclear pMad signal level . Data are plotted as nuclear pMad levels normalized to that of the co-processed control specimens . Image analyses were performed in Adobe Photoshop . Wandering third instar larvae expressing mCD8::GFP in nociceptors under the control of ppk1 . 9-GAL4 were anesthetized by submersion in a drop of glycerol in a chamber that contained a cotton ball soaked by a few drops of ether . ddaC neurons in segments A4-6 were imaged on Zeiss LSM 5 Live with a 40x/1 . 3 Plan-Neofluar oil immersion objective lens . A series of tiled images were captured and assembled to reconstruct the entire dendritic field of the three A4-6 ddaC neurons . Z-stack images were then converted to maximum intensity projections . Dendritic field coverage was quantified as described previously [16] . A ppk1 . 9-GAL4; UAS>CD2 stop>mCD8::GFP hs-flp strain was used to induce single cell flip-out clones in order to sparsely label nociceptors . Six virgin females and three males were used to seed vials containing a cornmeal molasses medium for a period of 2 days . The seeded vials were then heat-shocked in a 35°C water bath for 30 minutes . After an additional 3 to 5 days , wandering third instar larvae were harvested from the vials and dissected . In order to precisely identify the neurons responsible for the axons labeled in the CNS , the incision made in filleting the larvae was along the dorsal side , and the CNS remained attached to the fillet prep during immunostaining . mCD8::GFP and rat CD2 were detected using rabbit anti-GFP and mouse anti-rat CD2 primary antibodies , and visualized by anti-rabbit Alexa488 and anti-mouse Alexa568 secondary antibodies , respectively . Axon terminal branches of single cell flip-out clones were imaged in the abdominal ganglion using a Zeiss LSM 5 Live with a 40x/1 . 3 Plan-Neofluar oil immersion objective . The cell body of origin for each flip-out clone was then determined by inspecting the body wall of the corresponding fillet . Flip-out clones belonging to A1-7 segments were imaged and analyzed . To analyze the projection patterns for axon terminals , the presence or absence of terminal branches in each neuromere and longitudinal tract was manually identified for each single nociceptor clone . In order to align clones projecting to different segments , positions relative to the entry neuromere were used . The neurons that aligned were then used to calculate the percentage projecting to each neuromere and longitudinal tract . Heat-maps were color-coded according to these percentages using Microsoft Excel and Adobe Illustrator . The quantification of axon terminal area was performed in Matlab . Z-stack images of axon termini were converted to maximum intensity projections and manually cropped to exclude signals from other clones in the same sample . The green channel ( GFP ) and red channel ( CD2 ) of the cropped images were separately binarized using Otsu’s method [70] . The number of GFP-positive pixels were counted to calculate the area innervating the termini . To compensate for differences in the size and shape of the ventral nerve cord , the number of GFP-positive pixels was normalized to the average size of a single neuromere , which was calculated as the number of CD2-positive pixels divided by the number of neuromeres in the cropped image . To analyze axon terminals in nociceptors after 24-hour tkvQD expression ( see below ) , GFP and CD2 signals were linearly enhanced to match to the control images in order to compensate low expression level of GFP and CD2 . The clones whose signal intensities were too low to be binarized by Otsu’s method were excluded from the analysis . Larvae raised in normal fly vials for 5 or 6 days at 25°C , or larvae raised on apple juice plates containing ATR for 4 days at 25°C , were transferred to 30°C for 24 hours . In every experiment , experimental genotypes and control animals were treated side-by-side to minimize the effect of potential variations in temperature . The ppk1 . 9-GAL4 UAS-GCaMP6m strain was crossed to either a control strain ( w1118 ) or UAS-tkvQD strain . Activity of larval nociceptors were monitored at their axon terminals in the larval ventral nerve cord ( VNC ) , which was exposed for imaging by a partial dissection as follows: wandering third instar larvae expressing GCaMP6m in Class IV md neurons were immobilized in ice cold hemolymph-like saline 3 . 1 ( HL3 . 1 ) ( 70 mM NaCl , 5mM KCl , 1 . 5 mM CaCl2 , 4 mM MgCl2 , 10 mM NaHCO3 , 5 mM Trehalose , 115 mM Sucrose , and 5 mM HEPES , pH 7 . 2 ) [71] . The outer cuticle of each larvae was cut at segment A2/A3 to expose the central nervous system from which intact ventral nerves innervate the posterior larval body . The partially dissected animals were transferred to an imaging chamber containing HL3 . 1 equilibrated to the room temperature ( 23–25 °C ) . A strip of parafilm was placed over the larval VNC and was used to gently press the nerve cord down onto a coverslip for imaging . A Zeiss LSM5 Live confocal microscope and a 20x/0 . 8 Plan-Apochromat objective with a piezo focus drive were used to perform three-dimensional time-lapse imaging . Z-stacks consisting of 10–11 optical slices ( Z depth of 63 to 70 μm ) of 256 x 128 pixel images were acquired at approximately 4 Hz . During imaging , and using a custom-made thermal probe , a heat ramp stimulus was applied locally to one side of the A5 to A7 segments . The temperature of the thermal probe was regulated using a variac transformer . 10V was used to generate a 0 . 1 °C/sec heat ramp stimulation and no voltage was applied during cooling . A thermocouple probe ( T-type ) wire was placed inside of the thermal probe to monitor the probe temperature , and the data were acquired at 4 Hz through a digitizer USB-TC01 ( National Instruments ) and NI Signal Express software ( National Instruments ) . The acquired images and temperature data were analyzed using Matlab software ( Mathworks ) . Maximum intensity projections were generated from the time-series Z-stacks . Region of interest ( ROI ) was selected as a circular area with a diameter of 6 pixels , whose center was defined as the centroid of the A6 neuromere . Averaged fluorescent intensities ( F ) were calculated for the ROI for each time point . The average of Fs from the first 30 frames was used as a baseline ( F0 ) , and the percent change in fluorescent intensity from baseline ( ΔF/F0 ) was calculated for each time point . Since acquisitions of images and probe temperatures were not synchronized , probe temperature for each time point was estimated by a linear interpolation from the raw probe temperature reading . For a comparison of controls and tkvQD OE , ΔF/F0 , data were binned and averaged in 1°C intervals . To statistically compare proportional data , Fisher’s exact test was used . Multiple comparisons of proportional data were corrected by the Bonferroni method . For non-proportional data , Mann-Whitney’s U-test was used for pair-wise comparisons , and Steel’s test ( non-parametric equivalent of Dunnet’s test ) was used for multiple comparisons . Statistical analyses were performed in R software and Kyplot . | Although pain is a universally experienced sensation that has a significant impact on human lives and society , the molecular mechanisms of pain remain poorly understood . Elucidating these mechanisms is particularly important to gaining insight into the clinical development of currently incurable chronic pain diseases . Taking an advantage of the powerful genetic model organism Drosophila melanogaster ( fruit flies ) , we unveil the Highwire-BMP signaling pathway as a novel molecular pathway that regulates the sensitivity of nociceptive sensory neurons . Highwire and the molecular components of the BMP signaling pathway are known to be widely conserved among animal phyla , from nematode worms to humans . Since abnormal sensitivity of nociceptive sensory neurons can play a critical role in the development of chronic pain conditions , a deeper understanding of the regulation of nociceptor sensitivity has the potential to advance effective therapeutic strategies to treat difficult pain conditions . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"medicine",
"and",
"health",
"sciences",
"optogenetics",
"nociception",
"immunology",
"social",
"sciences",
"neuroscience",
"motor",
"neurons",
"clinical",
"medicine",
"nociceptors",
"brain",
"mapping",
"nerve",
"fibers",
"bioassays",
"and",
"physiological",
"analysis",
"hypersensitivity",
"research",
"and",
"analysis",
"methods",
"sensory",
"physiology",
"animal",
"cells",
"axons",
"sensory",
"receptors",
"neurophysiological",
"analysis",
"signal",
"transduction",
"psychology",
"cellular",
"neuroscience",
"cell",
"biology",
"clinical",
"immunology",
"physiology",
"neurons",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"sensory",
"perception",
"cell",
"signaling",
"bmp",
"signaling"
] | 2018 | BMP signaling downstream of the Highwire E3 ligase sensitizes nociceptors |
Molecular signaling networks are ubiquitous across life and likely evolved to allow organisms to sense and respond to environmental change in dynamic environments . Few examples exist regarding the dispensability of signaling networks , and it remains unclear whether they are an essential feature of a highly adapted biological system . Here , we show that signaling network function carries a fitness cost in yeast evolving in a constant environment . We performed whole-genome , whole-population Illumina sequencing on replicate evolution experiments and find the major theme of adaptive evolution in a constant environment is the disruption of signaling networks responsible for regulating the response to environmental perturbations . Over half of all identified mutations occurred in three major signaling networks that regulate growth control: glucose signaling , Ras/cAMP/PKA and HOG . This results in a loss of environmental sensitivity that is reproducible across experiments . However , adaptive clones show reduced viability under starvation conditions , demonstrating an evolutionary tradeoff . These mutations are beneficial in an environment with a constant and predictable nutrient supply , likely because they result in constitutive growth , but reduce fitness in an environment where nutrient supply is not constant . Our results are a clear example of the myopic nature of evolution: a loss of environmental sensitivity in a constant environment is adaptive in the short term , but maladaptive should the environment change .
Adaptive evolution optimizes the fitness of organisms for their environment through the accumulation of beneficial mutations by natural selection [1] . While we understand much about the mechanisms by which natural selection operates , less is known about the beneficial mutation rate [2] , and the genetic basis of adaptation [3] . Of particular interest is the spectrum of mutations that are adaptive in a specific environment , defined here as “adaptive strategy” . Through the use of experimental evolution , in conjunction with technological innovations such as candidate gene sequencing [4]–[8] , cDNA , [9] , [10] and tiling microarrays [11] , [12] , and whole genome sequencing of individual clones [13]–[16] and populations [17] , [18] , the field has recently gained significant insight into the genetic basis of adaptation . However , while candidate gene sequencing is certainly incomplete ( though still instructive ) in the picture it provides , even the identification of all mutations in individual clones does not reveal a complete representation of adaptation . Sequencing a handful of selected clones from an experiment provides only a microcosm of the adaptive mutational spectrum , while sequencing many clones from an experiment begins to resample the most prevalent lineages . By contrast , sequencing terminal clones from many different experiments ( e . g . [14] ) provides deeper insight into the convergence or divergence of the adaptive process , but is unable to capture evolution in action , including the clonal interference that occurs in the typically large populations used in microbial experimental evolution systems . To capture the dynamics of the adaptive process , as well as the mutational spectrum that accompanies it , it is necessary to sequence very large numbers of clones , possibly from many time points during an experiment , or instead to sequence entire populations as they evolve . In large asexual populations , selection acts positively to increase the frequency of the lineages containing beneficial mutations , while competition between coexisting adaptive lineages reduces the overall rate at which beneficial mutation increase in allele frequency , a process termed clonal interference [19] , [20] . Clonal interference occurs when beneficial mutations are sufficiently common to allow multiple adaptive lineages to expand in the population concurrently [21] . By deeply sequencing populations at multiple time points it is possible to not only identify mutations , but to also track the evolutionary dynamics of adapting lineages . Three studies published thus far have performed whole genome sequencing of evolving populations [17] , [18] , [22] , identifying SNPs at as low as 5% allele frequency in the sequenced populations; in the first two of these studies , E . coli were evolved by serial transfer , effectively in a continuously varying environment . The second of these two studies [18] , sequenced deeply enough that the allele frequencies of identified mutations over time could be tracked . However , it is likely that at ever-lower allele frequencies , there will be more observable beneficial mutations , most probably with smaller fitness effects . In the third of the studies , 40 replicate yeast populations were propagated by serial transfer for 1 , 000 generations , and sequenced every 80 or so generations [22] , allowing allele frequencies to be able to be determined down to 10% allele frequency . To better enumerate the adaptive strategy under a particular environment and to gain a better quantitative measure of the extent of clonal interference , deeper sequencing is needed however , which will likely identify additional mutations at lower allele frequencies with which to better characterize the adaptive mutational spectrum . Different environments likely result in different adaptive strategies , and many natural environments are variable and unpredictable , with irregular fluctuations in environmental parameters . Consequently , signaling networks evolved to enable organisms to be able to sense and respond to uncertain environments [23] . Signaling networks are ubiquitous across the Tree of Life , yet the question remains , “are functional signaling networks an essential feature of a well-adapted biological system ? ” Intracellular symbionts have undergone extensive genome reductions , likely due to relaxed selection in a setting that has few environmental perturbations . A major functional theme in these genome reductions is the loss of genes involved in signaling and genetic regulation [24] , [25] . However , this loss is likely neutral gene degradation due to genetic drift rather than adaptive evolutionary processes [26] . We sought to determine if the loss of environmental sensitivity is a viable or indeed preferred adaptive strategy . A constant environment provides an opportunity for such a system to evolve , since environmental sensing is superfluous , and perhaps even carries a fitness cost . We characterized the adaptive strategy , and the dynamics of adaptive lineages of the budding yeast S . cerevisiae evolving in a constant environment by ultra deep genome- and population-wide sequencing of three parallel evolution experiments .
Our population sequencing shows that clonal interference plays a dominant role in all three experiments , as 74 of the identified mutations ( 63% ) decrease in frequency following their maxima , and 42 of these mutations ( 57% ) become extinct by the end of the experiment ( Figure 2 ) . These results agree with theoretical predictions [19] , [28] and previous observations [29] , and imply that even if a mutation rises to a level above our detection threshold , it is still likely to succumb to an expanding fitter lineage and eventually become extinct . Evolution under conditions that promote clonal interference is also predicted to promote the accumulation of multiple beneficial mutations within an adaptive lineage before the first mutation can sweep [21] . We genotyped clones to determine the linkage of mutations above 10% frequency , and find that 91% of these mutations coexist in a clone with one or more other mutations . This value is an underestimate , since most mutations ( 67% ) never reach 10% frequency and thus were not analyzed for linkage . While from sequencing data alone we cannot unequivocally label a mutation as beneficial versus neutral , recurrent independent mutations ( see below and Figure S2 ) are likely to be beneficial . By this definition , all lineages that we were able to define by genotyping carry at least 1 beneficial mutation ( Figure 3 ) . Furthermore , the “winning” lineages occupying the largest proportion of the final population carry at least three beneficial mutations , and at least five mutations total ( Figure 3 ) . An exceptional case is E1 , where six mutations occur in close succession ( four of which are genes that we observe as recurrently mutated ) and result in what appears to be a complete selective sweep ( Figure 3a ) . These data indicate that multiple beneficial mutations – often occurring in close succession on what appears to have been a wild-type background – are necessary for a lineage to be successful . However , having multiple mutations is not sufficient for a lineage's success; for example , three lineages in E1 , each with two mutations , become extinct due another lineage sweeping ( Figure 3a ) . Furthermore , almost two thirds ( 49/76 ) of recurrent , and thus likely beneficial mutations never reach 10% frequency . The dynamics of adaptation suggest the “survival of the luckiest” , where for a new beneficial mutation to reach a high frequency , it must occur on a background that already has multiple other beneficial mutations [29] . This makes predicting the outcome of adaptive evolution difficult since the fixation probability of a beneficial mutation is no longer deterministic and proportional to the selection coefficient , but is also dependent on the genetic background on which the mutation occurs , which is distinctly a chance event . Our data show unequivocally that clonal interference between lineages carrying multiple beneficial mutations defines the dynamics of adaptation . We sought to understand the adaptive strategy of yeast growing in a constant environment by categorizing the genes in which mutations had occurred . Grouping recurrently mutated genes by pathway , we find that 53% of these mutations across all experiments reside in genes which function in three major signaling pathways: glucose signaling and transport , cyclic adenosine monophosphate/protein kinase A ( cAMP/PKA ) and the high osmolarity glycerol ( HOG ) response pathway ( Figure 4 ) , and these pathways have statistically enriched GO terms ( Table S3 ) . To further characterize the adaptive strategy , we characterized mutations by their predicted consequences . We found that the majority ( 73% ) of mutations are predicted to disrupt protein function , with nonsense mutations being enriched by 7 . 6-fold ( p<2 . 2e-16 ) ( Figure 5 ) . Together , these data suggest that the general adaptive strategy in a constant environment is the loss of signal transduction pathway function ( Figure 4 ) . For the glucose signaling pathway , disruptive mutations in MTH1 and RGT1 lead to constitutive expression of the glucose transporter ( HXT ) genes [30] , [31] , which increases the amount of glucose that is able to enter the cell , facilitating growth and providing a selective advantage [27] . The cAMP/PKA pathway positively responds to glucose in wild-type cells leading to growth [32]; disruptive mutations in the three recurrently mutated repressors GPB2 , IRA2 and PDE2 would cause constitutive pathway activation and growth , while loss of function in RIM15 ( the second most mutated gene , with 7 mutations observed ) , which is repressed by the PKA pathway , is akin to having increased PKA activity through that downstream path . Rim15 function is involved in the establishment of stationary phase [33] – presumably loss of the ability to enter stationary phase must be beneficial in the constant chemostat environment . The HOG pathway mediates transcriptional response to hyper-osmotic stress and also causes a brief growth arrest [34] , so the observed disruptive mutations in pathway activators would be expected to eliminate this response . All five HOG pathway mutations we genotyped occur in lineages with pre-existing MTH1 or RGT1 mutations , ( Figure 3a–b ) , suggesting potential positive epistasis between the HOG and glucose signaling pathways . To assess the extent of parallel adaptation we examined the overlap in genes and GO terms between experiments . E1 , E2 and E3 share 50% , 61% and 21% of their mutated genes with one of the other experiments , with E1 and E2 having the most sharing . MTH1 , RIM15 and GPB2 are mutated in all three experiments , with MTH1 being the most frequently observed mutated gene , having 19 independent mutations observed . We grouped enriched GO terms that share edges into GO networks to eliminate redundant GO terms and determined that E1 and E2 share all GO networks with each other ( Table S4 ) . E3 has overlap with the other two experiments , with 3 of 6 networks shared with both E1 and E2 . The GO network overlap suggests that the replicate experiments followed similar functional trajectories , with the underlying mutations broadly impacting similar biological processes in all experiments , namely the disruption of environmental sensing and signal transduction . We have shown that loss of environmental sensing through disruptive mutations in signaling pathways is adaptive in a constant environment . As signaling pathways make organisms robust to environmental changes , we hypothesized this loss would have a fitness cost in environments where nutrient availability was not constant . We thus subjected 18 clones containing mutations in one or more signaling pathway to starvation conditions . All 18 clones lost viability more rapidly than wild-type ( Figure 6a ) . To understand which mutations were causing decreased viability , we assayed nine strains containing single mutations from E3 [27] for premature cell death , and found that mutations in or downstream the cAMP/PKA pathway ( 4 of 9 mutations assayed ) showed significantly lower cell viability during starvation ( Figure 6b ) . Of these , we have previously shown that mutations in 3 of these 4 genes are beneficial alone in a glucose limited chemostat [27] . Thus , our results suggest that the adaptive strategy utilized by yeast in the constant chemostat environment is maladaptive in an environment where nutrients are not constant , indicating that there is an evolutionary trade-off due to antagonistic pleiotropy ( e . g . see [35] ) .
We have previously used fluorescent markers to track subpopulations during adaptive evolution in a constant environment [12] , and observed clonal interference in each of the 8 experimental populations that we evolved , in concordance with theoretical expectations . In this work , we have greatly expanded upon this , by performing ultra deep whole genome , whole population sequencing at each of 8 timepoints across 3 of these experimental evolutions . In addition to allowing us to identify mutations at an allele frequency as low as 1% , these population sequence data also provide us with direct estimates of the frequency trajectories of the mutations . Of the 3 experiments , only one resulted in a fixation event ( E1 , where 4 mutations in the same lineage were fixed by the final time point at 448 generations ) . By contrast , most mutations that enter the population were at a lower frequency than their maximum by the end of the experiment , and indeed more than a third had gone extinct . In most cases , mutations that were subject to clonal interference were in genes that were recurrently mutated ( 53 out of 74 ( 72% ) ) , and of those mutations that went extinct , the majority were also in genes that were recurrently mutated ( 32 out of 42 ( 76% ) ) . Thus , clonal interference clearly plays a major role in these populations in shaping their eventual composition , with many beneficial mutations in the population going extinct during the evolution . A recent study which also used sequencing of populations undergoing experimental evolution [18] did not observe such a great extent of clonal interference , though in their experiments they only could detect mutations that reached a 5% allele frequency . In our data , of the 74 mutations we detected that were at a lower frequency by the end of the experiment than their maximum frequency ( i . e . were subject to clonal interference ) , 38 ( 51% ) had a maximum frequency of less than 5% . Thus , deeper sequencing is able to provide significantly more insight into the process of clonal interference . We observed 35 mutations in genes that were recurrently mutated that failed to reach a 5% frequency in the experiment , though we only identified 2 additional recurrently mutated genes by being able to get to allele frequencies lower than 5% ( OSH3 and LCB3 ) . There were 19 mutations that did not reach an allele frequency of 5% that were in genes that were not recurrently mutated – further experimentation to determine whether these mutations are adaptive , and/or even deeper sequencing would be required to confidently extend the adaptive mutational spectrum . We also observed that multiple mutations prevail , with all of the lineages that we detect as present in our populations at the end of the experiment carrying more than one mutation , with at least two predicted to be beneficial . It is an open question as to how many lineages with beneficial mutations actually existed within the population – there are few empirical estimates of the beneficial mutation rate , and those that do exist are based on a relatively modest number of observed mutations . One estimate , based on mutations that fixed in Pseudomonas fluoresecens , is 3 . 8e-8 per cell division [36] . If that were similar to the beneficial mutation rate in yeast , then with a population of 1e9 growing for 448 generations , we might expect as many as 17 , 000 beneficial mutations to occur within any one of our experiments . Most of these would not be expected to establish – if we assumed that ∼10% establish ( roughly similar to an average 10% fitness effect ) , then 1 , 700 lineages with beneficial mutations would have established in a given experiment . By contrast , Shaw et al [37] , analyzing mutation accumulation lines in A . thaliana , found that approximately half of all mutations observed were beneficial . In yeast , also using mutation accumulation lines , Hall and colleagues have estimated that between 5% and 13% of mutations are beneficial [38]–[40] . With a per base pair mutation rate on the order of ∼1e-10 [41] and a genome size of 12e6 , the number of cells estimated per generation to have a mutation is around 1 in 1 , 000 . If 10% of mutations are beneficial , then 1 in 10 , 000 per generation may receive a beneficial mutation . Thus , in our experiments , we might expect as many as 50 million beneficial mutations to occur over the 448 generation time course , with ∼5 million establishing . While these are estimates based on relatively small number of mutations in mutation accumulation lines , even if they are over estimated by 2 orders of magnitude , it is clear that sequencing of even hundreds of randomly selected individual clones ( which will likely represent a few , prevalent lineages ) , or even deep population sequencing will not be able to fully characterize the spectrum of beneficial mutations , nor determine an accurate estimate of their fitness effects . While to our knowledge this study is the deepest sequencing yet performed on experimentally evolving populations , it may only represent the tip of the potential adaptive iceberg ( though this is likely the most important part , as these mutations likely drive the evolutionary process ) , while our previous work [12] was only the tip of the tip . New , higher throughput approaches , and rational ways of identifying and selecting independent lineages are clearly needed to fully understand this most fundamental of biological processes . We observed the parallel evolution of mutations that disrupt one or more of three major signaling pathways responsible for sensing environmental stimuli and responding by governing growth rate . We propose a model for the adaptive strategy in constant , nutrient-limited environments ( or at least in a glucose limited environment ) ( Figure 7 ) , wherein constitutive commitment to cell division is beneficial , and thus mutations that result in unrestrained cell division are adaptive as long as the growth rate does not exceed the influx rate of nutrients into the system . By and large , these mutations are loss of function mutations . We consider the mutations in these pathways to be decoupling the sense and response to environmental stimuli , leading to an adaptive loss of environmental sensing in a constant environment . In contrast , these mutations are maladaptive in environments where nutrient abundance is not constant , such as when going through a boom and bust cycle from high glucose into starvation conditions . This may be due to depletion of the cell's reserve nutrient supply or the inability to enter a quiescent state , leading to cellular death . This adaptive loss of environmental sensitivity is a powerful example of how evolution is myopic: by evolving strategies to cope with a constant and predictable environment , genes and pathways are disrupted that would be necessary for survival when cells are confronted with an uncertain environment . It is noteworthy that the clones characterized in Wenger et al [35] , also evolved in an aerobic glucose limited environment , were also more fit under a diverse set of other carbon limited environments , suggesting that their adaptive strategy also translated to other constant environments . Whether this strategy is widely applicable under an array of constant environments with different nutrient limitations remains to be determined through additional experimentation . However , recent analysis of experiments in bacteria have verified the idea that loss of function mutations can be a general strategy for adaptive evolution [42] . From a broad viewpoint , the adaptive strategy of loss of environmental sensitivity that we observed is similar to the strategy tumor cells use to proliferate . Cancer is an evolutionary process of clonal selection [43]–[45] , and it is beneficial for the cells to replicate as fast as possible through the accumulation of mutations in oncogenes and tumor suppressor genes , many of which are in the homologous Ras/cAMP/PKA pathway that is recurrently mutated in our experiments [46] , [47] . While the external environment humans face is dynamic and unpredictable , the human body has evolved to maintain homeostasis , exemplified by the near constant concentration of blood glucose [48] . Such mutations also come with trade-offs – when faced with an uncertain environment , many of the mutations show antagonistic pleiotropy ( AP ) . In our data , three single mutants that we tested had a reduced fitness in the starvation environment , for which we have previously demonstrated fitness gains in the chemostat environment where they were selected – clearly cases of AP . For the multiple mutants that we tested , the loss of fitness could be due to AP , or alternatively result from a prior hitchhiking event in the evolved environment of a mutation that is deleterious in the starved environment . This mutation accumulation hypothesis ( MA ) is considered as an alternative to antagonistic pleiotropy when an evolved lineage shows fitness trade-offs . The fact the many , if not all of the mutations in our multiple mutants are in genes or pathways that are recurrently mutated in our chemostat evolutions makes MA seem a less likely explanation than AP . Indeed , previous experiments using E . coli evolving by serial transfer [49] showed that the rate of loss of unused functions in parallel evolving populations was consistent with AP , rather than MA , suggesting that AP may be widespread , and that when evolving in a consistent ( though not necessarily constant ) environment that due to the fitness cost of unneeded pathways , that there is a use it or lose if effect [50] . It has been shown using the yeast deletion collection that AP is indeed widespread , with approximately 20% of the collection of non-essential gene deletions being more fit under one of the tested conditions [51] . It is also of note that in evolving E . coli strains , mutations that result in a loss function of the sigma factor encoded by rpoS ( which is involved in the general stress response ) are frequently selected [52] . These mutations frequently exhibit AP , being detrimental under conditions where there is stress , the response to which needs to be balanced with growth ( see [53] for review ) . Most of the AP examples thus provided are loss of function mutations ( either from systematic gene deletion projects , or from sequencing beneficial mutations arising during experimental evolution ) , but a systematic catalog of AP effects of large numbers of beneficial mutations has not yet been generated . It will also be interesting to determine how clones with beneficial mutations that exhibit AP can perform adaptive escape when allowed to evolve afresh in an environment in which their previously beneficial mutations are now deleterious .
All population samples in this study have been previously described [12] . Briefly , three strains of haploid S288c that are isogenic , except that each constitutively expresses a different fluorescent protein ( GFP , YFP or DsRed ) , were seeded in equal quantities in a 20 mL chemostat device . Each population was evolved for 448 generations at steady state under glucose limitation ( 0 . 08% ) at a dilution rate of 0 . 2 h−1 . During this evolution , the proportions of the three colored lineages were tracked using flow cytometry , and population samples were archived under deep freeze in glycerol at −80°C at regular intervals . Wild-type ancestral strain GSY1136 was also used as a reference for sequencing . Illumina sequencing libraries were made directly from glycerol stocks of the original population samples , as well as the wild-type ancestral strain ( GSY1136 ) . Stocks were melted , and genomic DNA was extracted from 500 µl of each stock using Zymo Yeast Genomic DNA columns . The Nextera library prep kit ( Epicentre ) was used to construct the libraries , starting with 25–50 ng of genomic DNA . The tagmentation reaction was performed in LMW Reaction Buffer at 55°C for 10 minutes . The resulting tagged DNA was subjected to PCR using the Nextera PCR enzyme ( Epicentre ) under the following conditions: 72°C for 3 min , 95°C for 30 sec; 9 cycles of 95°C for 10 sec , 62°C for 30 sec , 72°C for 10 sec; final extension at 72°C for 1 min . A shortened extension time was used to bias the amplification of short fragments in order to maximize the proportion of bases being sequenced twice with overlapping paired-end Illumina reads . A modified Adapter 2 with a random hexamer barcode of sequence 5′-CAAGCAGAAGACGGCATACGAGATNNNNNNCGGTCTGCCTTGCCAGCCCGCTCAG-3′ ( PAGE-purified , IDT Technologies ) was used during the PCR for the population samples , while the standard Nextera Adapter 2 was used for GSY1136 . No size selection was performed on the libraries , although they were concentrated through a Qiagen MinElute column . The same GSY1136 library was spiked into all 24 population libraries at a molar rate of 5% . The resulting libraries were sequenced on one lane apiece of 2×101 bp plus a 6 bp index read on the Illumina Hi-Seq 2000 . In addition , two independent libraries from the same genomic DNA of GSY1136 were sequenced on one Hi-Seq lane apiece . An overview of the sequencing analysis pipeline used to identify variants is given in Figure S3 . The wild-type GSY1136 library that was spiked into each population sample was extracted with the exact tag ATCTCG using a modified version of the Fastx Barcode Splitter ( http://hannonlab . cshl . edu/fastx_toolkit/index . html ) . Nextera adapters were trimmed off the 3′ read ends with Cutadapt v0 . 9 . 4 [54] supplied with the Nextera adapter sequence and default parameters except -m 15 . The resulting FASTQ files were culled of any reads that occurred in only one read of the pair . Paired-end reads were mapped to a custom S288c reference genome with BWA ( bwa-short ) v0 . 5 . 9-r16 [55] using default parameters plus -I -q 10 , and a sorted BAM file was created with Picard v1 . 45 FixMateInformation ( http://picard . sourceforge . net ) . The custom genome was constructed as follows: single end Illumina reads of a different ancestral wild-type strain ( GSY1135 ) from a previous study [27] were mapped to the S288c reference sequences from the Saccharomyces Genome Database ( SGD; http://www . yeastgenome . org/; downloaded 2/24/2011 ) . SNPs were called with the GATK v1 . 0 . 5777 UnifiedGenotyper [56] , [57] , and a FASTA reference sequence was constructed that incorporated these SNP calls using the GATK FastaAlternateReferenceMaker . The population data were culled of PCR duplicates using a modified version of Picard MarkDuplicates . In this program , the random hexamer barcodes were used in addition to the mapping coordinates to decide if a pair of reads was a PCR duplicate . Specifically , if more than one read pair had the same mapping coordinates in addition to the same hexamer barcode , only the pair with the highest mapping quality was retained for further analysis . The in-lane spike-in of wild-type library was used to recalibrate the base qualities of the population data from the same lane . To achieve this , GATK CountCovariates and TableRecalibration were called on each lane of the wild-type data separately , using a variant mask for the CountCovariates step created by Samtools v0 . 1 . 16 [58] mpileup . Recalibration was visualized as successful as visualized by AnalyzeCovariates . The covariate file from the wild-type recalibration was then used as input for TableRecalibration on the population data from the same lane . Proper recalibration was assessed once again by AnalyzeCovariates . A custom Java program was written to identify the bases in each library fragment that were sequenced twice by overlapping read pairs . This analysis was applied to both population and wild-type data , and the overlap information was stored in the custom “ZO” tag of the BAM file . A Python script implementing PySAM v0 . 5 ( http://code . google . com/p/pysam/ ) was used to calculate the allele counts for each position in the reference genome , and the following filters were applied: uniquely mapping reads only , base quality score greater than 19 required , and only bases sequenced twice that were concordant in base identity between the two reads were retained . Population SNP calls were made by comparing the allele counts in each population sample for each genomic position to the counts of the same allele and position from the wild-type data . This comparative approach filtered out any position that had false positive SNP calls due to positional effects , such as mapping or systematic sequencing errors . First , a merged wild-type file was created by combining all the spike-in control wild-type data with the two independently sequenced wild-type files . Second , only non-reference alleles that had both an allele count of at least 2 and a larger frequency in the population sample than the wild-type were retained . Third , a one-tailed Fisher's Exact Test was used to calculate if the number of non-reference alleles out of all alleles at a site was significantly greater in the population data than in the wild-type data for the same allele . These p-values were FDR corrected using the Benjamini and Hochberg method [59] , and only sites with a q-value less than 0 . 01 were retained . The following heuristic post-hoc filters were applied to the set of SNPs: 1 ) SNPs with a maximum frequency that was greater than the largest color proportion , plus 0 . 1 , for the appropriate time point were removed ( color frequency data from [12] ) . This removes any SNP that rose to a higher frequency than the highest color , which is not possible , unless identical SNPs arose in different colored populations . 2 ) Any SNP that was significant in the first time point was removed . This is because even if a new mutation present at the start of the experiment conferred a relative fitness of 2 , that mutation would not be detectable in our assay in the first sampled generation . 3 ) Any site that was not deemed callable was removed . Callability was determined empirically with the GATK CallableLociWalker ( -frlmq 0 . 01 -minMappingQuality 2 ) on the relevant population data , as well as the merged wild-type data . 4 ) Sites that had greater than 5% non-reference alleles in the merged wild-type data were removed . These sites were largely systematic errors . 5 ) SNPs where the read position of the variant allele did not vary were removed . This was defined as a read position standard deviation lower than one . 6 ) SNPs that had a mapping quality bias between reads containing the reference and variant alleles were removed , as calculated by a Bonferroni-corrected Mann-Whitney U test on mapping qualities . Mutation allele frequencies were validated against a set of known mutation frequencies for experiment C1 ( Figure S4 ) with data from [12] , [27] , as well as the fluorescent protein reporter frequencies for all experiments ( Figure S5 ) . All putative SNPs with a maximum allele frequency greater than 10% were confirmed by Sanger sequencing , except for the chr16:581589 mutation in experiment E2 , which we were unable to amplify by PCR . While no effort was made to comprehensively catalog indels , Sanger sequencing of putative SNPs revealed six indels , which in every case were due to mapping errors of true indels near the ends of reads . Co-occurrence of SNPs was determined by Sanger sequencing clones picked from the relevant time points for mutations greater than 10% allele frequency . The effect of each SNP ( non-coding , synonymous coding , non-synonymous coding , etc . ) was established with SNPeff v2 . 0 . 3 ( http://snpeff . sourceforge . net/ ) . The permissiveness of all missense mutations was calculated using SIFT [60] with default parameters . To create the lineage dynamics plots , allele frequency data were plotted assuming linear expansion or contraction between primary data points . Since the allele frequency data were of lower resolution than the flow cytometry data ( 8 vs 47 time points ) , sometimes the inferred linear extrapolation between frequency data points resulted in an allele frequency greater than the color frequency . In these cases , the extrapolated allele frequencies were reduced to fit within the bounds of the color frequencies . Note , this fitting was performed for extrapolated points only; primary allele frequency data remained untouched . All mutations discovered across the three experiments were divided into the following coding mutation effect categories: stop gained , start lost , stop lost , non-synonymous and synonymous . The sum of mutations within these categories was compared to the expectation using a chi-squared test . The expectation was calculated empirically by assuming random mutation throughout the genome; i . e . all possible mutations in the genome were made in silico , and the effect of the mutation was assigned to one of the categories above . The expected proportion of each category was calculated as the total for each category out of all possible mutations , and this proportion was multiplied by the total number of mutations discovered to get the expected number of mutations for each category . The same analysis was performed for coding versus non-coding mutations . To find an enrichment of disruptive versus tolerated mutations , the totals of the stop gained , start lost , stop lost and disruptive non-synonymous categories were summed into the “disruptive” meta-category , and the synonymous , tolerated non-synonymous and non-coding mutations were summed into the “tolerated” meta-category . The SIFT predictions were used to classify non-synonymous mutations as either disruptive or tolerated . Expectations for disruptive or tolerated non-synonymous mutations were calculated empirically by summing the SIFT effect of all possible mutations for a particular protein . Cell viability was quantified under starvation conditions using propidium iodide ( PI ) and flow cytometry similar to [61] , in biological triplicate . Overnight cultures in 1 . 2 mL YPD were grown unshaken in deep-well 96 well plates at 30°C . Cultures were spun down , aspirated , and resuspended in 1 . 2 mL sterile water , and then diluted 1∶3 into a minimal medium described previously [12] supplemented with 2% glucose . The cultures were left undisturbed at 30°C between time points . Cell viability was measured at regular intervals post-inoculation by mixing the cultures and diluting 50 µL of culture into 250 µL water containing 250 µg PI , following by analysis by flow cytometry . The proportion of viable cells was calculated as PI-negative cells out of total cells analyzed . Significantly different viability was calculated with a two-tailed t-test between each mutant strain and wild-type at each time point . Cell viability based on PI staining was validated by colony forming unit analysis on a subset of the strains analyzed . Gene Ontology ( GO ) biological process enrichments of coding mutations for each experiment were calculated using GO::TermFinder [62] at SGD with default options except “Feature Type” set to “ORF” and dubious ORFs disqualified from the analysis . For the reproducibility analysis , GO terms sharing edges were grouped into networks and GO networks were considered shared between experiments if they had at least one shared GO term . All Illumina sequencing data are available from the NCBI Sequence Read Archive with accession number SRA054922 . | When a population of organisms is faced with a selective pressure , such as a limiting nutrient , mutations that arise randomly may confer a fitness benefit on the individual carrying that mutation . If that individual reproduces before it is lost from the population , the frequency of that mutation may increase . Over time , many beneficial mutations will arise in a large population , but there are few high resolution experiments tracking the frequency of such mutations in an evolving population . We evolved populations of the baker's yeast in a constant environment in the presence of limiting amounts of sugar , and then used DNA sequencing to identify mutations that reached at least a 1% frequency in the population , and tracked them over time . We identified 120 mutations over three experiments , and determined that the genes and pathways that had gained beneficial mutations were largely reproducible across experiments , and that many of the mutations led to the loss of signaling pathways that usually sense a changing environment , allowing the cell to respond appropriately . When these mutant cells were faced with uncertain environments , the mutations proved to be deleterious . Environmental sensing must carry a fitness cost in a constant environment , but is essential in a changing one . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2013 | Whole Genome, Whole Population Sequencing Reveals That Loss of Signaling Networks Is the Major Adaptive Strategy in a Constant Environment |
The International HapMap project has made publicly available extensive genotypic data on a number of lymphoblastoid cell lines ( LCLs ) . Building on this resource , many research groups have generated a large amount of phenotypic data on these cell lines to facilitate genetic studies of disease risk or drug response . However , one problem that may reduce the usefulness of these resources is the biological noise inherent to cellular phenotypes . We developed a novel method , termed Mixed Effects Model Averaging ( MEM ) , which pools data from multiple sources and generates an intrinsic cellular growth rate phenotype . This intrinsic growth rate was estimated for each of over 500 HapMap cell lines . We then examined the association of this intrinsic growth rate with gene expression levels and found that almost 30% ( 2 , 967 out of 10 , 748 ) of the genes tested were significant with FDR less than 10% . We probed further to demonstrate evidence of a genetic effect on intrinsic growth rate by determining a significant enrichment in growth-associated genes among genes targeted by top growth-associated SNPs ( as eQTLs ) . The estimated intrinsic growth rate as well as the strength of the association with genetic variants and gene expression traits are made publicly available through a cell-based pharmacogenomics database , PACdb . This resource should enable researchers to explore the mediating effects of proliferation rate on other phenotypes .
The International HapMap project [1] , [2] has made available a vast amount of genetic variation data from a large number of individuals with diverse ethnic background . A recent population based whole-genome sequencing initiative ( 1000 Genomes Project [3] ) sought to expand on this effort by providing a more comprehensive catalog of human genome sequence variation , including rare variants in these samples . These data can be used to study the effect of genetic variants on disease processes , pharmacologic traits , and environmental responses . As part of the HapMap project , EBV-transformed lymphoblastoid cell lines ( LCLs ) derived from individuals of diverse ancestry were established , which provide renewable sources of DNA and RNA . The commercial availability of these cell lines and the rich genetic information publicly available have enabled a large number of researchers to adopt them as in vitro models for the study of genotype-phenotype relationships in human cells [4] . Consistent with this trend , a vast amount of phenotypic data such as gene expression levels , drug response , and radiation response have been made publicly available [5]–[8] . Furthermore , an enormous amount of genotype-phenotype relationships have been generated [4] , [9]–[11] . Our group has therefore constructed a database , PACdb [6] , a public central repository of pharmacology-related phenotypes , to host these integrative results obtained in HapMap LCLs . Although there are many advantages in utilizing the cell-based system for genotype-phenotype studies , the problem of biological and experimental noise when dealing with LCL-based phenotypes and the potential for spurious results has been recognized by several researchers [12] , [13] . Indeed , it has been proposed [14] that non-genetic confounders and other technical factors in generating phenotypes from these cell lines may hamper efforts to evaluate the genetic contributions to phenotype . One common factor , cellular growth rate , has undergone scrutiny for its effect on various phenotypes , particularly drug-induced cytotoxicity , a phenotype of interest in pharmacogenomic studies [12] , [13] . For example , aberrant growth rate is one of the distinctive features of cancer cells and growth inhibition following exposure to chemotherapeutics and other cytotoxics is intimately related to growth rate [13] . Thus , studying cellular proliferation rate is likely to advance our understanding of cancer pathogenesis . In this study , we set out to extract and combine data from various sources and calculate intrinsic cellular growth rate using a novel mixed effects model ( MEM ) for over 500 HapMap cell lines . Previous studies have shown the presence of a strong correlation between gene expression traits and growth rate in other organisms such as yeast and bacteria [15]–[17] . Brauer et al . [15] measured gene expression traits in yeast under several controlled growth conditions and reported that 25% of the gene expression phenotypes were correlated with growth rate . In addition , genes important for cellular proliferation have been found to be differentially expressed in most cancer tissues [18]–[20] . Such genes were shown to be strong prognostic factors in breast cancer [21]–[23] , renal cancer [21] , lung cancer [21] , mantle cell lymphoma patients [24] . Thus , to gain insights into the factors contributing to intrinsic growth rate phenotype , we also evaluated the relationship between gene expression and cellular growth .
We have computed the intrinsic growth rate of these cell lines using MEM ( described in Materials and Methods section ) and provide the values in Table S1 . Figure 1 shows these values in comparison with the raw data . The rightmost boxplot in each panel ( label 12 ) represents the intrinsic growth rate . The other boxplots correspond to the raw data . It is clear that the variability of the raw data is in general much larger than the intrinsic growth . Unless otherwise stated , all subsequent analyses are done on the intrinsic growth rate . The mean and standard deviation of the intrinsic growth by population is shown in Table 1 . The effect of population on growth rate was significant ( likelihood ratio test ) with ASW growing the fastest , followed by YRI and ASN , and CEU being the slowest . In pairwise comparison , only CEU's growth was significantly ( ) lower than ASW's . Similar results had been reported for YRI , ASN and CEU by Stark et al . [13] but the wide range of experimental conditions used in the current study strengthens the evidence . Figure 2 shows boxplots of intrinsic growth rate by population . Since we have reduced the effect of confounders by combining data from multiple experiments , the differences we find are likely to be intrinsic to the cell lines . However , the generalizability of these results to in vivo population differences is not obvious . One factor that may explain the slow growth of CEU cell lines is the time from the establishment of these lines and merits further study . The mean and standard deviation of the intrinsic growth rate by gender are also shown in Table 1 . Figure 3 shows the boxplot of the latter by gender and population . The effect of gender on growth rate was found to be significant . Across all experiments , female cell lines grew approximately 7% ( 95% CI ) slower than male cell lines . This finding remained quite consistent across individual experiments . Gender differences are not likely to be due to experimental differences since both female and male cell lines are handled similarly . Differential effect of estrogen and other hormones in media could explain part of this difference . Choy et al . [14] measured growth rate on some of the same cell lines and made the data publicly available ( CEU I/II , YRI I/II , and ASN ) . They performed cell counting for five consecutive days and estimated growth rate as the slope of the fitted line ( log concentration vs . time ) . This growth rate is based on only one biological replicate ( although they did measure an additional biological replicate at one time point for a subset of cell lines for internal validation ) . We compared our estimated intrinsic growth rate with the growth rate measured by Choy et al . [14] . The correlation between Choy's growth rate and ours was 0 . 30 , which supports the idea that both our intrinsic growth and Choy's growth are realizations of the true intrinsic growth rate albeit with different degree of noise . We analyzed Choy's data and found comparable magnitude of the effect of gender ( slightly less than 4%; p = 0 . 07 ) and cell lines from CEU population grew at the slowest pace in both datasets . Baseline gene expression data for CEU and YRI phase I and II described in Zhang et al . [27] were used to examine the association with the intrinsic growth rate . We found that almost 3000 out of the 10748 genes examined were associated with the intrinsic growth rate at FDR0 . 10; the result held regardless of whether we adjusted for expression heterogeneity or not [28] . The list of genes whose expression level associated with intrinsic growth rate in CEU and YRI LCLs is provided in Table S2 . Figure 4 shows the QQ-plot of p-values from the association between gene expression phenotypes and intrinsic growth rate adjusted by gender and population . The upper left panel shows the QQ-plot for the unadjusted analysis and the upper right panel shows the plot for the Surrogate Variable Analysis ( SVA; expression heterogeneity ) adjusted analysis . The lower panels show the corresponding histograms with p-values highly concentrated near the zero . The gray dots in each of the QQ-plots correspond to p-values under the null hypothesis of no association , which was computed by randomly permuting the phenotype . Actual p-values lie well above the null hypothesis p-values , which indicate that the significance found is not due to model misspecification or correlation between gene expression phenotypes . Stark et al . [13] had not been able to find significant association ( FDR0 . 10 ) between gene expression and growth rate in the first phase CEU population because of the noisier version of growth rate and smaller sample size used at the time . The strong correlation between gene expression traits and cellular growth rate is consistent with similar findings in yeast [15] and in bacteria [16] , [17] . A clear advantage of using our method is shown by the fact that our power to detect association between gene expression phenotype and growth rate is substantially increased . This is illustrated in Figure 5 , which compares the p-values from the association between gene expression phenotype and growth rate when the intrinsic value computed with MEM is used vs . when the individual experiment's values are used . The left panel shows the p-values from the SVA ( expression heterogeneity ) adjusted analysis and the right panel shows the results from the unadjusted analysis . All points lie below the one-to-one line , which means that the intrinsic growth rate achieves greater power in identifying association than any of the individual experiment's data . The red dots correspond to growth rate data from Choy et al . [14] . Functional enrichment analysis was performed using DAVID Bioinformatics Resources [29] , [30] . Table 2 shows the GO terms that were enriched in our intrinsic growth rate-associated gene set . They clustered into cell cycle , cell death , intracellular transport , protein transport and phosphorylation . For comparison , we checked the proportion of cell cycle , mitosis , cell death and phosphorylation genes among metabolic process genes ( from GO ) and immune response genes ( from GO ) . The proportion of growth-associated genes annotated to these terms were 6 . 8% , 2 . 7% , 6 . 1% , and 6 . 2% , respectively . In comparison , among immune response genes 0 . 2% , 0 . 16% , 1 . 1% , and 0% were annotated to these terms . For metabolic process genes none of these terms reached significance at the loose threshold of . Table 3 shows the SP-PIR keywords enriched in our growth gene set; more than half of the growth-associated genes were associated with phosphoprotein ( p ) and 22% of them were associated with acetylation ( p ) . For comparison , we checked the proportion of genes related to phosphoprotein keyword for two other cellular functions: metabolism ( metabolic process from GO ) and immune response . None of these genes were annotated with the phosphoprotein keyword in the SP-PIR database . Table S3 shows the Kegg pathways enriched in our set . We compared the growth-associated genes with two recently published proliferation signatures . The first one was obtained by performing a meta-analysis of over 2833 breast tumor expression profiles by Wirapati et al . [31] . The second one was compiled by Starmans et al . [21] based on cell cycle in cervix cancer cell lines [32] and human fibroblasts [33] . We found that 44% of Wirapati's proliferation genes belonged [31] to our growth-associated gene list ( defined as ) and 75% of them had a positive effect on growth ( higher expression associated with faster growth ) . This enrichment is not likely to occur by chance as can be seen in Figure 6 . The figure shows a histogram of the number of growth-associated genes we would get if we randomly sampled the set of all genes we considered . The vertical line indicates the actual number of growth-associated genes in Wirapati's list . We performed the same analysis with Starmans et al . [21] proliferation signature but did not find any significant enrichment . Nevertheless , 80% of the Starmans' proliferation genes had a positive effect on LCL growth rate . We performed a genome wide association study ( GWAS ) of the intrinsic growth rate for CEU and YRI unrelated cell lines . Even though we were unable to find genome-wide significant SNPs , we did find that the top intrinsic growth-associated SNPs were more likely to target ( as eQTLs ) intrinsic growth-associated genes . We have quantified the enrichment using three different procedures described in the Methods section . The first one uses the hypergeometric distribution to test the enrichment of growth associated genes among targets of growth associated SNPs . The Fisher's test p-value was . The second method accounts for the correlation between genes and yielded an empirical ( none of the 1000 simulations yielded Fisher's test p-value smaller than the observed ) . The third method accounts for the fact that the data from the same individuals are used when computing the intrinsic growth-gene association as well as the intrinsic growth-SNP association . The empirical p-value with this method was 0 . 026 . Since eQTLs have been shown to be more likely to be associated with complex phenotypes [34]–[37] , we focused the analysis on growth-associated gene eQTLs but the enrichment was not strong enough to render significant SNPs after adjusting for multiple testing . Interestingly however , among the growth-gene eQTLs we found two well-replicated colorectal cancer SNPs: rs4779584 [38] , [39] and rs3802842 [39] , [40] . A target gene of rs4779584 is a growth-associated gene NEU1 [MIM:608272] ( growth gene expression association ) , which has been reported to contribute to the suppression of metastasis of human colon cancer [41] . A target gene of rs3802842 , MED13 [MIM: 603808] ( growth gene association q-value = 0 . 052 ) , is part of the CDK8 subcomplex [42] and CDK8 is a colorectal oncogene that regulates beta catenin activity [43] . To our knowledge , the potential functional connection between these two established colorectal cancer SNPs and the growth-associated genes NEU1 and MED13 has not been made previously . The association between these SNPs and intrinsic growth rate however was not significant . As an attempt to address the concern of whether these associations may be confounded by EBV transformation , we cross-checked the growth-associated genes with the list of EBV transformation-associated genes reported by Caliskan et al . [44] and found no evidence of enrichment of EBV genes among our growth genes . Furthermore , we analyzed the effect of EBV copy numbers on intrinsic growth . For this purpose , we used measurements of EBV copy numbers on a large portion of cells used for the association with gene expression ( 86 CEU , 74 YRI ) from Choy et al . [14] . We found a small but significant effect ( ) of EBV on intrinsic growth . However , once we accounted for SVA variables ( expression heterogeneity ) , the effect was no longer significant . Similar results were found using EBV data generated in our lab for a subset of the samples . This result suggests that the intrinsic growth-gene expression associations we found ( SVA adjusted ) are not mediated by EBV copy numbers , consistent with the lack of enrichment in EBV-related genes among our top growth genes .
In this study , we propose a novel method , MEM , that combines data from multiple sources using a mixed effects model and estimates an intrinsic phenotype that is more reflective of the true phenotype than each of the individual experiment's data . We apply it to generate intrinsic cellular growth rate , which is a phenotype with important implications for disease biology and phamacogenomics . Using MEM we computed the intrinsic growth rate of over 500 HapMap cell lines and studied their properties . To our knowledge , this is the most comprehensive analysis to date of intrinsic cellular growth rate for the HapMap cell lines , for which various biological conditions were included in the estimation . We understand that estimates of intrinsic growth can be further improved as more experiments are included . A Bayesian approach would fit well for this purpose . Existing data would make up the prior distribution for the intrinsic growth rates and the addition of new data would generate posterior distributions , presumably more concentrated on the true intrinsic growth rates . We plan to regularly update the HapMap LCL intrinsic growth rate phenotype data and make them widely available to the research community through PACdb [6] . We found significant in vitro population differences in cellular growth rate in the HapMap populations included in our study . The ASW lines ( African American ) proliferated at the fastest rate followed by YRI and ASN ( Asian ) , and CEU were the slowest . We analyzed Choy et al . 's [14] growth rate data and also found CEU lines to grow slower than other populations . Since we combined data from multiple sources and reduced the level of noise , the observed population differences are likely to be intrinsic to the cell lines and may in part be due to genetic factors; however , the methods used in establishing the LCLs and the experimental conditions during the EBV-transformation could also contribute to this observation . The fact that CEU cell lines were established much earlier than other populations could in part explain their slow growth . Of the populations included in the HapMap Project , only the CEU LCLs existed as previously established cell lines . The other populations were collected and established as cell lines specifically for the HapMap Project over the years 2002 through 2007 [12] , [13] . Nonetheless , the observed population difference in intrinsic cellular growth rate needs to be considered when studying population differences in complex traits using these cell lines . Interestingly , we found that female cell lines grow at roughly 7% slower pace than male cell lines consistently across different experiments . We found similar gender effect when we analyzed Choy et al . 's data [14] . Gender differences are not likely to be due to experimental differences since both female and male cell lines are handled similarly . Differential effect of estrogen and other hormones in media could explain part of this difference . It is not clear whether these observed population and gender differences are extensible beyond these cell lines . However , these initial observations warrant further studies . We found that almost 3000 gene expression phenotypes were associated with the intrinsic growth rate , which is consistent with findings in yeast [15] . This finding held robustly , whether or not we accounted for expression heterogeneity using Surrogate Variable Analysis [28] . Our top growth-related genes were enriched in cell cycle , mitosis , cell death , and phosphorylation terms . We also found a significant overlap between our intrinsic growth genes and a proliferation signature inferred from breast tumor microarray data [31] . Thus , our study provides a comprehensive list - combining both germline and tumor cells - of potential biomarkers and therapeutic targets for proliferation-mediated phenotypes . Furthermore , the gene expression traits associated with intrinsic growth determined by our study are much more significant than their corresponding associations with growth rate data generated from any individual experiment , including Choy et al . 's [14] growth rate . This strongly suggests that our method to combine data from several experiments is succeeding at yielding a more intrinsic measure of growth rate . Despite the limited power given the relatively small sample size used for eQTL mapping , we demonstrate evidence of genetic effect on intrinsic growth rate by determining the enrichment of growth-associated genes among genes targeted by top growth-associated SNPs ( as eQTLs ) after accounting for LD structure and correlation between gene expressions . Interestingly , among intrinsic growth gene eQTLs , we found two well replicated colorectal cancer SNPs ( rs4779584 [38] , [39] and rs3802842 [39] , [40] ) , which target growth-associated genes NEU1 and MED13; both genes have been implicated in colorectal cancer [41]–[43] . We deposited our findings into PACdb [6] , which should be a useful addition to the already rich set of phenotype data currently available for the HapMap cell lines . In addition to the intrinsic growth rates ( Table S1 ) , the significance of the association with gene expression phenotypes ( Table S2 ) is available from the same database . This resource should be useful to explore any mediation effect of growth rate on the phenotype of interest either by using the intrinsic growth rate as a covariate in the analysis or by looking at overlap between the phenotype of interest and the top growth related genes . We also make the R code to apply MEM and generate intrinsic growth available on PACdb ( http://pacdb . org/growthrate/generate-igrowth . r and http://pacdb . org/growthrate/rawgrowth . txt )
MEM pools phenotype data from multiple sources and computes an intrinsic value of the phenotype for each individual after accounting for different experimental conditions and covariates . where the index identifies the individual or cell line , i represents the intrinsic phenotype of the individual , experimental conditions can represent a large range of different experimental conditions ( for example different sites , technicians , method , passage number , etc . ) , 's are relevant covariates , and is an error term . The index represents different replications of the data for given individual and experimental condition . The intrinsic phenotype and experimental conditions are treated as random effects and covariates 's are treated as fixed effects . Extension to generalized linear model is straightforward . HapMap cell lines were purchased from the non-profit Coriell Institute for Medical Research ( http://www . coriell . org/ ) and growth rates were measured using alamarBlue assay as described in Stark et al . [12] . The alamarBlue assay gives a measure of the number of proliferating cells at time :where is some increasing function . Thus growth rate can in principle be computed aswhich is also some function of the alamar number . With a slight abuse of notation , we will refer to this alamar number as the growth rate . It should be noted that the approach we describe here holds generally regardless of the assay used to measure the number of proliferating cells . We use MEM to compute an intrinsic growth rate for each cell with the following model:where the index identifies the cell line , pop is the population to which the cell line belongs to ( CEU , YRI , ASW , or ASN ) , iGrowth0 represents the intrinsic growth rate of the cell line , experimental conditions can represent a large range of different experimental conditions , and is an error term . The index represents different replications of the data for given individual and experimental condition . We set the gender and population as fixed effects and experimental conditions and iGrowth0 as random effects . The term iGrowth0 is a cell line specific intercept ( so it is different for each cell line ) and can be interpreted as ( some monotone function of ) the intrinsic growth rate of each individual . This term is ( by construction ) orthogonal to gender and population . In general , it may make more sense to include the population and gender effect in the intrinsic growth rate so we define iGrowth as the sum of the iGrowth0 and the estimated effects of population and gender . The term “experimental conditions” could be allowed to be more than one-dimensional . For our dataset it was sufficient to use “technician” as the experimental condition . The reason for this choice was that each technician's work was for the most part concentrated at roughly the same time ( within 6 months ) so the experimental conditions such as thaw history are likely to be reasonably homogeneous . Our results were robust to using other combinations of experimental conditions such as a combination of technician , drug and population . We found no need to account for the trio structure since the correlation coefficient between parent and child was not significantly different from zero . This fact should not be interpreted as lack of heritability but that the level of noise was too high to be able to estimate the correlation with the given sample size . Growth rate itself was quite normally distributed . The additivity assumption of the model may be better achieved in the log scale but we did not notice much difference in the overall results when we tried different transformations so we used the untransformed variable . We fit the mixed effects model using the lme4 [45] package for the R Statistical Software [46] . P-values for the fixed effects were calculated using likelihood ratio tests after fitting the models with maximum likelihood option ( REML = FALSE ) . Gene expression data was generated by our lab for phase I/II CEU and YRI cell lines using Affymetrix GeneChip Human Exon 1 . 0 ST array as described by Zhang et al . [27] . Association between gene expression and growth rate was calculated using a linear model with log-transformed gene expression data as response and intrinsic growth , population and gender as covariates . Surrogate Variable adjustment was done using the SVA package [28] , [47] . FDR was computed using Storey's qvalue package[48] , [49] . Figures were generated using the graphic capabilities of R and the ggplot2 package [50] in R . Functional term enrichment was assessed using DAVID [29] , [30] . Genes associated with intrinsic growth at % were used as significant genes and the default Homo Sapiens list was used as background . Genome wide association between genotype and intrinsic growth was performed using the PLINK v1 . 97 software [51] ( http://pngu . mgh . harvard . edu/purcell/plink/ ) . CEU ( I/II and II ) and YRI ( I/II and III ) unrelated cell lines were used with draft release 2 consensus genotype ( which passed QC across all 11 populations from HapMap 3 samples ) downloaded from the HapMap Project website . We use three methods to assess the enrichment of growth associated genes among targets of growth associated SNPs . First we use the hypergeometric distribution and tests ( Fisher's test ) whether the overlap is more significant than one would get with a random set of genes . This method computes the exact p-value but assumes independence between genes . The second method accounts for the correlation structure between target genes by simulation , which will induce a correlation structure between simulated genes similar to the observed one . For this purpose , we permute the phenotype 1000 times and for each permutation we perform GWAS , select the top SNPs ( ) , query the target genes for the top SNPs ( ) using SCANdb , and calculate the Fisher's test p-value for the overlap between growth genes and target genes . Finally , we compute an empirical p-value for the enrichment as the proportion of times the simulated Fisher's p-value was smaller than the observed Fisher's p-value . The third method accounts for the fact that the data from the same individuals are used when computing the intrinsic growth-gene association as well as the intrinsic growth-SNP association . The simulated target genes are generated as described in the second method . The intrinsic growth associated genes are generated using the same permuted phenotype used to generate the target genes ( FDR ) . The association p-value is computed by regressing the simulated intrinsic growth on gene expressions ( without expression heterogeneity adjustment , ) . For each simulation , the Fisher's p-value is computed for the overlap between simulated target genes and simulated intrinsic growth associated genes . Finally , we compute an empirical p-value for the enrichment as the proportion of times the simulated p-value was smaller than the observed p-value . | Cell-based models provide a convenient system to conduct studies that would be impossible to apply to human subjects , but the phenotypes measured on these models can be marred with biological noise . We propose a method ( MEM ) to address this issue by statistically combining data from various sources , and we apply it to the proliferation rates of cell lines collected as part of the International HapMap project . We show that the proliferation rate computed using our method is a better measure of the true proliferation rate of the cells and produces a much stronger association with gene expression phenotypes on the same cell lines: more than 30% of the genes tested were significantly associated with proliferation rate . We also demonstrate that genetic variants have an effect on growth rate . Finally , we make these intrinsic proliferation rates and the strength of the association with gene expression phenotypes public , which should allow other researchers to explore the mediating effects of proliferation on other phenotypes . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"genomics",
"genetics",
"biology",
"computational",
"biology",
"genetics",
"and",
"genomics"
] | 2012 | Mixed Effects Modeling of Proliferation Rates in Cell-Based Models: Consequence for Pharmacogenomics and Cancer |
The APOBEC3 deoxycytidine deaminase family functions as host restriction factors that can block replication of Vif ( virus infectivity factor ) deficient HIV-1 virions to differing degrees by deaminating cytosines to uracils in single-stranded ( − ) HIV-1 DNA . Upon replication of the ( − ) DNA to ( + ) DNA , the HIV-1 reverse transcriptase incorporates adenines opposite the uracils , thereby inducing C/G→T/A mutations that can functionally inactivate HIV-1 . Although both APOBEC3F and APOBEC3G are expressed in cell types HIV-1 infects and are suppressed by Vif , there has been no prior biochemical analysis of APOBEC3F , in contrast to APOBEC3G . Using synthetic DNA substrates , we characterized APOBEC3F and found that similar to APOBEC3G; it is a processive enzyme and can deaminate at least two cytosines in a single enzyme-substrate encounter . However , APOBEC3F scanning movement is distinct from APOBEC3G , and relies on jumping rather than both jumping and sliding . APOBEC3F jumping movements were also different from APOBEC3G . The lack of sliding movement from APOBEC3F is due to an 190NPM192 motif , since insertion of this motif into APOBEC3G decreases its sliding movements . The APOBEC3G NPM mutant induced significantly less mutations in comparison to wild-type APOBEC3G in an in vitro model HIV-1 replication assay and single-cycle infectivity assay , indicating that differences in DNA scanning were relevant to restriction of HIV-1 . Conversely , mutation of the APOBEC3F 191Pro to 191Gly enables APOBEC3F sliding movements to occur . Although APOBEC3F 190NGM192 could slide , the enzyme did not induce more mutagenesis than wild-type APOBEC3F , demonstrating that the unique jumping mechanism of APOBEC3F abrogates the influence of sliding on mutagenesis . Overall , we demonstrate key differences in the impact of APOBEC3F- and APOBEC3G-induced mutagenesis on HIV-1 that supports a model in which both the processive DNA scanning mechanism and preferred deamination motif ( APOBEC3F , 5′TTC; APOBEC3G 5′CCC ) influences the mutagenic and gene inactivation potential of an APOBEC3 enzyme .
APOBEC3F ( A3F ) and APOBEC3G ( A3G ) are members of a family of seven single-stranded ( ss ) DNA cytosine deaminases ( A3A , A3B , A3C , A3D , A3F , A3G , and A3H ) [1] and play a role in restriction of the retrovirus HIV-1 ( referred to as HIV ) [2] . Research has been highly focused on primarily A3G and secondarily A3F for a number of years since they appeared to be the most efficient restrictors of HIV replication [3] , [4] , [5] , [6] , [7] . Although there are documented restrictive effects of A3G , and possibly A3F , at an individual level ( reviewed in [8] ) , the suppression of HIV by A3G and A3F at a population level is lost due to the HIV protein Vif ( viral infectivity factor ) [6] , [9] . Vif forms an E3 ubiquitin ligase with host proteins and causes A3G and A3F polyubiquitination and degradation through the proteasome [6] , [10] , [11] , [12] , [13] , [14] . The general mechanism by which A3G restricts HIV , which has been a paradigm for other A3 enzymes , requires that it be encapsidated with the ribonucleoprotein complex of HIV [9] , [15] . A3G requires its N-terminal domain ( NTD ) , which can bind nucleic acids , for encapsidation into virions [16] . A3G catalyzes deaminations through its C-terminal deaminase domain ( CTD ) [16] , [17] . In the target cells that these virions infect , encapsidated A3G can deaminate cytosines to uracils in ( − ) DNA reverse transcribed from the RNA genome , after the reverse transcriptase associated RNaseH activity enables ssDNA regions on the ( − ) DNA to be accessed by the enzyme [18] , [19] . The uracils in the ( − ) DNA are used as a template by reverse transcriptase during ( + ) DNA synthesis and result in guanine to adenine mutations . If A3G can induce sufficient numbers of these mutations , the resulting proviral DNA will be functionally inactivated . The deaminases A3D , A3F , and A3H appear to follow this general mechanism of restriction in cell culture , but to differing degrees than A3G [3] , [4] , [20] , [21] , [22] , [23] , [24] . The exceptions are A3A , which inhibits incoming HIV viral particles in myeloid lineage cells [25] , [26] , A3C , which does not appear to become encapsidated or restrict HIV in cell culture [20] , [27] , and A3B , which can restrict HIV in 293T and HeLa cells , but not SupT1 cells [20] , [28] . Despite a possible role for A3F , A3D and A3H Haplotype II in HIV restriction , it appears that A3G is more effective at restricting HIV replication and that perhaps the other A3 enzymes function in a collaborative way with A3G [20] , [21] , [22] , [23] , [24] , [29] . In particular there has been a recent focus on the restriction capability of A3F . A3F was initially identified as potentially being an equal contributor with A3G to the restriction of HIV [3] , [5] , [6] , [7] , [30] , but current research demonstrates , in agreement with an earlier report [4] , that A3F may have less antiviral activity than A3G [22] , [23] , [24] , [29] . Many different experimental protocols , such as analysis of stably expressed A3F from a cell line [23] , use of primary cell lines [24] , and A3F haplotypes from donor samples [22] have been applied and demonstrate that A3F has less of an effect on HIV infectivity in comparison to A3G . However , another report showed no difference in restriction efficiency of A3G and A3F beyond 2-fold using experiments that knocked-down endogenous A3 expression in a nonpermissive cell line [21] . As a result , the role of A3F in restriction of HIV remains unclear . Among reports demonstrating less of an affect of A3F in restricting HIV replication than A3G , there is still no identified reason for why this may occur . From some reports A3F mRNA is expressed 10-fold [31] or 5-fold [22] less than A3G mRNA , suggesting less A3F would become virion encapsidated . However another report found A3F and A3G mRNA expression was more comparable [32] . Further , some reports have found a direct correlation with mRNA and protein levels [31] , [32] whereas other reports have been unable to make such a correlation due to the use of different primary antibodies [24] . Confounding the interpretation of these data are reports which demonstrated that A3F is preferentially encapsidated with the HIV ribonucleoprotein complex in comparison to A3G [4] , [33] . Song et al . concluded that the encapsidation difference between A3G and A3F in effect absolves any difference in cellular expression [33] . Despite this observed more specific packaging of A3F in the ribonucleoprotein complex [33] , studies have found a minimal contribution of A3F to the hypermutation of HIV genomes or less potency in HIV-1 restriction [4] , [22] , [23] , [24] , [29] . Together these data suggest that if there is a difference in restriction efficiency of A3F and A3G , that it is not the physiological conditions which cause different effects on HIV infectivity , but an inherent difference in their biochemical characteristics . However , there has been no in depth biochemical characterization of A3F to date to determine what might be these differences between A3G and A3F . As such , we have undertaken a characterization of A3F in comparison to A3G to identify an underlying biochemical reason for these observations . In particular , we have focused on characterizing the mechanism A3F uses to scan ssDNA . This is because it has been shown that the ssDNA scanning mechanism of A3G is important for inducing mutagenesis of ( − ) DNA formed during reverse transcription of RNA [34] . A3G has been characterized to scan ssDNA through facilitated diffusion [35] , [36] , [37] . Facilitated diffusion is a 3-dimensional scan of DNA by enzymes to locate their target sites for catalysis [38] , [39] , [40] . The movement is characterized by sliding , jumping or intersegmental transfer motions . Sliding is used to describe short range 1-dimensional scanning motions and can enable an in depth search of a particular area of DNA for a target motif [39] , [40] . Jumping is a term that describes microdissociations of the enzyme from the DNA with a reassociation on the same DNA substrate , i . e . , the enzyme does not diffuse into the bulk solution [38] , [41] . The negative charge of the DNA establishes a charged radius around the DNA molecule in which a positively charged enzyme can dissociate , diffuse and still return back to the same DNA . These jumping events enable enzymes to translocate larger distances than sliding thus making the search of non-target DNA more efficient than sliding alone [39] , [40] . Intersegmental transfer is similar to jumping but describes a movement where an enzyme with two DNA binding domains interacts with two distal sites simultaneously before dissociating from one of the sites [39] , [40] . Different research groups , including our own , have found A3G to use a combined sliding and jumping search mechanism [34] , [35] , [37] , [42] , although one report found A3G to use intersegmental transfer [36] . We have characterized A3G mutants and A3G in complex with different Vif variants that resulted in decreases of either sliding or jumping motions and found that the ability of these A3G forms to induce mutagenesis of nascently reverse transcribed DNA was decreased [34] , [43] . We have hypothesized that both sliding and jumping are important for inducing mutagenesis because A3G needs to conduct local searches ( sliding ) to effectively deaminate many cytosines , ensuring gene inactivation , and also translocate ( jumping ) over RNA/DNA hybrids to reach distal regions of ( − ) DNA [34] . The processive scanning of other A3 enzymes has not been reported , except A3A which was found to be largely nonprocessive [44] . This work is the first biochemical characterization of A3F and provides a biochemical explanation for the lowered ability of A3F to inactivate HIV , as reported by numerous research groups [22] , [23] , [24] , [29] and within this report . We have found that A3F primarily uses jumping movements to scan ssDNA which is detrimental to its ability to cause numerous mutations on ( − ) DNA during reverse transcription . The target motif of A3G ( 5′CCC ) also appears to cause more inactivating mutations in the HIV protease ( prot ) than the target motif for A3F ( 5′TTC ) , adding another level of deficiency in HIV inactivation potential . All together our data provide a model for the specific biochemical properties required for efficient restriction of HIV by A3 deaminases .
The processive nature of A3G has been shown to be of importance for inducing mutagenesis of HIV ( − ) DNA in a model in vitro system [34] , [43] and in cell culture [45] . It is not known whether A3F is processive . Since multiple lines of evidence from independent labs have shown that the effect of A3F on HIV is different than A3G [4] , [22] , [23] , [24] , we sought to determine if there was an inherent biochemical difference between the two enzymes that could account for these observations . Specifically we determined if there was a difference in the processive scanning mechanisms of these two enzymes with processivity being defined as the ability to deaminate more than one cytosine on an ssDNA in a single-enzyme substrate encounter . Processivity was determined using different synthetic ssDNA substrates containing two deamination motifs separated by different distances , 5′TTC for A3F and 5′CCC for A3G . This strategy was used since with A3G we have found that closely spaced deamination motifs , i . e . , 5 to 15 nt are deaminated most efficiently through sliding motions and as the distance between deamination motifs increases a jumping motion facilitates processive deaminations [34] . The substrate usage was kept below 15% to ensure single-hit conditions were maintained , which means that each ssDNA was only encountered by an enzyme at most once during the reaction [46] . On a substrate with the target cytosines separated by 30 nt ( Figure 1A , sketch ) , A3F was able to catalyze processive deaminations . The processivity factor is a ratio of the frequency of double deaminations on a single substrate to the predicted frequency of double deaminations of a nonprocessive enzyme ( see Materials and Methods ) . Therefore , the processivity factor of 3 . 7 for A3F ( Figure 1A ) means that in a single enzyme-substrate encounter A3F was 3 . 7-fold more likely to catalyze a processive deamination than a nonprocessive deamination . On the cognate A3G substrate , A3G was 2-fold more likely than A3F to catalyze a processive deamination ( compare Figure 1A , A3F , processivity factor of 3 . 7 and A3G , processivity factor of 7 . 9 ) , suggesting that the processive mechanisms of A3F and A3G differ . In addition , we observed a difference in the ability of A3F and A3G to catalyze 5′-end biased deaminations . Where A3G has been found to prefer deaminations towards the 5′-end of ssDNA molecules due to a catalytic orientation specificity [35] , A3F had a minimal 5′-end bias ( Figure 1A , compare intensity of 5′C & 3′C bands for A3F and A3G ) . However , the presence or absence of a 5′-end bias does not influence the processivity calculation [47] . Since A3G has been found to use a dual sliding and jumping motion to scan ssDNA [34] , [35] , [37] , we investigated whether the difference between A3F and A3G was due to a difference in the contributions of sliding and jumping or a different mode of scanning , e . g . , intersegmental transfer . First , we investigated the sliding ability of A3F . We conducted deamination assays on ssDNA substrates with closely spaced deamination targets , since it has been shown that sliding motions increase the frequency of closely spaced deaminations occurring processively [34] . With cytosines 14- and 5-nt apart , A3F was unable to catalyze any detectable processive deaminations ( Figure 1B–C , A3F , absence of 5′C & 3′C band ) indicating that A3F does not use sliding motions to catalyze processive deaminations . Of note , outside of single hit conditions ( >15% substrate usage ) we detected the band corresponding to deamination of both 5′TTC motifs on an ssDNA ( 5′C & 3′C band ) , which demonstrated that multiple molecules of A3F were able to deaminate these substrates at both cytosine targets to near completion ( data not shown ) . In contrast , A3G was able to processively deaminate closely spaced residues under single-hit conditions by sliding ( Figure 1B–C , A3G , processivity factors of 4 . 6 and 3 . 5 ) . Since A3F was processive on the substrate with the target cytosines separated by 30 nt ( Figure 1A , A3F ) , but not on substrates with closely spaced deamination motifs ( Figure 1B–C , A3F ) , the data suggested that A3F may use jumping or intersegmental transfer to processively deaminate cytosines . To investigate this further we determined the processivity of A3F on an ssDNA with the target cytosines separated by 63 nt ( Figure 1D ) . On this substrate , A3F exhibited a processivity factor of 4 . 6 ( Figure 1D , A3F ) , which is higher than the processivity factor obtained on the substrate with the target cytosines separated by 30 nt ( Figure 1A , A3F , processivity factor of 3 . 7 ) . In contrast , A3G which can slide and jump [35] maintained a processivity factor of ∼8 ( compare Figure 1A and D , processivity factors ) . To confirm that we would observe only jumping or intersegmental transfer and not sliding motions , we annealed a 20 nt complementary DNA in between the target cytosines ( Figure 1E , sketch ) . The double-stranded DNA portion is not bound as tightly by A3F ( Figure S1A–B ) or A3G ( Figure S1C and [35] , [37] , [48] , [49] ) as ssDNA ( Table 1 ) and results in the assay conditions blocking the sliding portion of the scanning activity [35] , [47] . A3G was still processive on this substrate due to the ability to translocate on DNA in 3-dimensions by jumping , but we observed a ∼2-fold decrease in A3G processivity as compared to the analogous ssDNA substrate ( Figure 1D–E , compare A3G processivity factors ) . We interpret that the ∼2-fold decrease in A3G processivity is due to A3G molecules attempting to slide over the dsDNA which induces dissociation from the DNA substrate and diffusion into the bulk solution . For A3F we observed a 1 . 8-fold increase in the processivity factor when we annealed a 20 nt complementary DNA in between the target cytosines ( Figure 1D–E , compare A3F processivity factors ) , despite having a reduced binding to the double-stranded ( ds ) DNA portion ( Figure S1A–B ) . The double deaminations became so efficient that the 5′- and 3′-proximal cytosine deaminations were barely visible on the gel ( Figure 1E , A3F ) . A3F bound the 118 nt ssDNA substrate ( Figure 1D ) with an apparent Kd of 20 nM ( Table 1 ) , which is ∼7-fold lower than the apparent Kd of A3G ( Table 1 , Kd of 130 nM ) . This indicates that A3F is less likely to dissociate from an ssDNA substrate than A3G , but does not fully explain why we observed an increase in processivity by annealing a complementary DNA in between the target cytosines . Results were not changed by annealing a 20 nt complementary RNA molecule to the substrate ( Figure S2A–B ) or by testing A3F on a different partially dsDNA substrate which contained two 5′ATC motifs ( Figure S3 ) . We speculated that the processivity of A3F increased as opposed to remaining the same in the presence of the complementary DNA because the structural change in the substrate induced by the dsDNA region made jumping events more successful . This could occur if the average jumping distance of A3F were different than A3G and the rigid dsDNA region juxtaposed the 5′TTC motifs at a distance which was highly accessible by A3F . To test this hypothesis we examined the processive deaminations of A3F and A3G on an ssDNA substrate with deamination motifs separated by 100 nt . We found that as the distance between deamination motifs was increased up to 100 nt , the processivity factors of A3F also increased ( Figure 2A ) . In contrast , A3G processivity exhibited a plateau when deamination motifs were 30- to 63-nt apart and the processivity factor decreased when deamination motifs were 100 nt apart ( Figure 2B ) . These data demonstrate that the average jumping distance of A3F and A3G differ . Similar results were also found from analysis of deamination-induced mutations in the model HIV replication assay and are discussed later in the text . To identify a possible reason for the different jumping ability of A3F we examined its oligomerization state in comparison to A3G . A3G is known to form polydisperse oligomers that are dependent on enzyme concentration and buffer conditions [50] . Using size exclusion chromatography at low enzyme concentrations we found that A3F formed predominantly tetramers ( ∼180 kDa ) and higher order oligomers whereas A3G eluted as predominantly a monomer ( ∼46 kDa ) with minor dimeric species ( Figure 2C ) . The finding that A3F forms more tetramers than A3G is consistent with previous sucrose gradient data [51] and data on the CTD portions of these enzymes . The A3F CTD can oligomerize more readily than the A3G CTD [52] , [53] , [54] , [55] . The A3F oligomers remained soluble as high speed centrifugation did not result in a discernable protein pellet . These data demonstrated that A3F oligomers are more stable than A3G oligomers at low protein concentration and suggest a structural difference that could account for why the A3F jumping distance is different than A3G ( Figure 2A–B ) . However , since the A3F ssDNA scanning mechanism is more efficient in distal translocations ( Figure 1 ) , we also investigated whether it was scanning ssDNA by intersegmental transfers , rather than or in addition to jumping . This mode of DNA scanning involves an enzyme molecule that binds in two distal locations on the DNA before completing the translocation by dissociating from one location [39] , [40] . The intersegmental transfer mechanism requires that the enzyme have more than one DNA binding domain . A3F could bind ssDNA with both its NTD and CTD on one or many subunits of the oligomer . This is in contrast to jumping which uses microdissociations and reassociations to scan ssDNA [38] , [41] . A key difference between jumping and intersegmental transfer is that the probability of an enzyme transferring to another DNA substrate is low for jumping but high for intersegmental transfer [39] , [40] , [56] . Therefore , to observe whether A3F can scan ssDNA by intersegmental transfer we increased the enzyme and substrate concentrations , but kept their ratio constant . Crowding the reaction in this manner with enzyme and ssDNA can increase the tendency of the enzyme to translocate to a different ssDNA if intersegmental transfer is occurring [56] . This would result in a decrease in the observed processivity with increasing reaction components . We found that A3F maintained the same processivity at a 1∶1 E∶S ratio at concentrations of 100 nM and 200 nM ( Figure 2D , processivity factor of 4 . 7 and 4 . 8 ) . At a 1∶1 E∶S ratio using concentrations of 300 nM and 400 nM the processivity of A3F decreased ∼1 . 5-fold from 4 . 7 to 3 . 0 or 3 . 3 ( Figure 2D ) , providing evidence that A3F can use intersegmental transfer to scan ssDNA . However , the decrease in A3F processivity is small ( ∼1 . 5 fold ) , does not decrease gradually with increasing enzyme and substrate concentration , and is not completely abolished ( processivity factor remains above 1 ) suggesting that intersegmental transfer is not the primary mechanism of DNA scanning , but can occur in a minority of ssDNA-A3F interactions . Importantly , intersegmental transfer should result in an increase in the reaction rate with increasing DNA concentration since the rate of searching is enhanced by increasing the apparent off rate , which allows more rapid sampling of DNA [56] . However , the reaction rate of A3F decreased with increasing enzyme and substrate concentrations ( Figure 2D , Rate ) and supports the conclusion that intersegmental transfer is not a primary mode of scanning ssDNA . In further support of this interpretation is that we only observed evidence of intersegmental transfer with increasing enzyme and substrate concentration ( Figure 2D ) , not when the ssDNA concentration alone was increased ( Figure 2E , processivity factors of 4 . 2 to 4 . 6 ) , which indicates that A3F does not readily transfer to another ssDNA without high local concentrations of enzyme , i . e . , the intersegmental transfer is not inherent to A3F but requires excessive crowding of reaction conditions . A3G showed no decrease in processivity with increasing concentration of enzyme and substrate , despite also containing both a NTD and CTD ( Figure S5 ) . This difference may arise since the CTD of A3G binds ssDNA in the micromolar range [54] , [55] , [57] , in contrast to the CTD of A3F that can bind DNA in the nanomolar range ( Table 1 , apparent Kd of 288 nM ) . All together the data supported the conclusion that A3F primarily utilized jumping and not intersegmental transfer to scan ssDNA . Our biochemical data on synthetic substrates ( Figure 1 ) predicts that A3F will not efficiently catalyze deaminations during proviral DNA synthesis due to a predominant jumping movement that would result in a superficial scan of the ssDNA [34] , [43] . Importantly , we observed this predominant jumping movement when A3F encountered an RNA/DNA hybrid ( Figure S2A ) , such as would be encountered during synthesis of the HIV provirus . To test this prediction we used our model in vitro HIV replication system . Since this system reconstitutes reverse transcription of ( − ) DNA and synthesis of ( + ) DNA , it allows us to observe the ability of A3 enzymes to induce mutagenesis in a dynamic system , such as occurs in vivo , but with the advantage of controlling the amount of enzyme added to the reaction system . Specifically , this system uses an in vitro synthesized RNA which contains ( from the 5′-end ) a polypurine tract ( PPT ) , part of the protease gene ( prot ) of HIV , and a lacZα reporter . The RNA is reverse transcribed to ( − ) DNA by reverse transcriptase and after RNaseH-mediated removal of the RNA , the PPT enables ( + ) DNA synthesis without the addition of an exogenous primer . In this manner we can achieve the salient properties of HIV replication that A3 enzymes must contend with , a finite time to access single-stranded ( − ) DNA and a heterogeneous substrate that is interspersed with RNA fragments . The A3G data demonstrated the potential amount of mutations that could occur in this system . A3G had a clonal mutation frequency of 2 . 63×10−2 mutations/bp which is 10-fold over the background mutation frequency of reverse transcriptase ( RT ) ( Table 2 ) . Further , the A3G mutation spectra have clear hot-spots at 5′CCC or 5′CC motifs in both the prot and lacZα with some sites being mutated in 100% of clones ( Figure 3A , e . g . , 245 nt ) . Due to the PPT being nearest the prot , this region is converted to dsDNA the fastest and incurs less mutations than regions nearer the center or 3′-end of the ( + ) DNA ( Figure 3A ) . As such , we can recover white colonies indicating a mutation in the lacZα reporter but upon sequencing find no mutations in the prot . Therefore , the number of clones with mutations in the prot is a measure of how efficiently an A3 enzyme can induce mutations . The lacZα remains single stranded longer and can therefore be visited by multiple A3 enzymes multiple times . In the prot region , A3G was found to induce no mutations in 13% of clones , but the majority of clones had either 1–2 mutations ( 47% ) or 3–4 mutations ( 37% ) ( Figure 3C ) . In the lacZα , A3G-induced mutagenesis resulted in >7 mutations in the majority of clones ( Figure 3D , 60% ) . Addition of A3F to the model HIV replication assay resulted in a modest 2 . 8-fold increase over the background mutation frequency ( Table 2 ) . Examination of the mutation spectrum demonstrated that A3F could induce mutagenesis at a number of 5′TTC or 5′TC sites along the prot and lacZα , but that there were no clear hot-spots , except possibly at position 305 nt of lacZα ( Figure 3B ) . This may be due to the random binding of A3F to the ( − ) DNA and an inefficient search of the enzyme by jumping without local scanning by sliding ( Figure 1 ) , which would make interaction with multiple 5′TTC or 5′TC motifs less likely to occur . Of note , the mutation frequencies induced by A3F and A3G did not increase with the addition of more enzyme to the reaction demonstrating that both A3F and A3G are present at saturating levels ( data not shown ) . Analysis of the distances between A3F-induced mutations demonstrated that 75% of the mutations were separated by more than 20 nt ( Table 3 ) , confirming that A3F was using jumping this assay system , in agreement with the data on the synthetic oligonucleotide substrates ( Figure 1 ) . In contrast , only 50% of A3G-induced mutations were separated by more than 20 nt ( Table 3 ) , providing confirmation that A3G is capable of recognizing sites that are more closely spaced ( Figure 2A–B ) . The analysis in Table 3 included all clones ( highly mutated and sparsely mutated ) . To ensure we did not bias our analysis we also examined only sparsely mutated clones for both A3G and A3F ( 2–5 mutations ) and obtained similar results for frequency of mutations separated by more than 20 nt ( A3G , 60% and A3F , 85% ) . In addition , we hypothesized that the tight binding of A3F to ssDNA ( Table 1 ) , would prevent A3F from frequently dissociating into the bulk solution and reassociating with different ( − ) DNAs . In agreement with the binding data , A3F increased the population mutation frequency ( frequency of white colonies ) only 9-fold over the background whereas A3G caused a 12-fold increase in the population mutation frequency ( Table 2 ) . Although the overall level of mutagenesis induced by A3F was low , we did observe slightly more mutations in the lacZα than the prot region due to the replication kinetics ( Figure 3E–F ) . In the majority of A3F clones ( 64% ) there were no mutations in the prot region ( Figure 3E ) . In the lacZα region the majority of clones only had 1–2 mutations ( Figure 3F , 36% ) . However , 29% of clones did not have a G→A mutation and were recovered due to an RT induced error ( Figure 3F , 0-0 ) . These data demonstrated that A3F was inefficient at inducing mutagenesis during reverse transcription even in areas where the enzyme had ample time to access ssDNA ( lacZα ) and especially in regions that are single-stranded the shortest time ( prot ) ( Figure 3E–F ) . The increased distance between A3F-induced mutation sites in the HIV replication assay ( Table 3 ) in combination with the data on synthetic oligonucleotide DNA indicating that A3F prefers to use jumping ( Figure 1 ) provides evidence that the decreased mutagenic ability observed for A3F in cell culture may be due to an inefficient search mechanism on DNA . However , these observations are inconsequential if each mutation by A3F were to inactivate the prot gene , which is used here as a predictor of HIV inactivation potential . We gauged the probability that the prot of HIV would be inactivated by A3F by determining the mutated amino acid sequences and comparing this to an extensive mutagenesis study of the prot conducted by Loeb et al . [58] . Consistent with A3F inducing a low number of mutations ( Figure 3 and Table 2 ) , there were no A3F-induced mutations in 64% of clones ( Figure 4B ) . On a per clone basis , A3F-induced mutations resulted in protease inactivation only 50% of the time ( Figure 4B , 18% active and 18% inactive ) . The high number of clones remaining active was due to two reasons . First , some clones incurred mutations in regions where any amino acid is tolerated [58] , even nonconservative changes , e . g . , E21K , so the mutation was insignificant ( Table S2 ) . Second , some clones incurred a mutation that resulted in a conservative change to the amino acid which enabled the protease to retain full or partial activity , depending on the proximity to the active site residues [58] ( Table S2 ) . For example , the M46I mutation was induced by A3F in 11% of clones , but results an active and drug resistant protease . The remainder of the A3F-induced mutations were found only in single clones and 36% of those mutations resulted in an active protease ( Table S2 ) . All together , A3F was not efficient at inactivating the HIV protease and could also induce resistance to protease inhibitors ( Table S2 , D30N and M46I ) . This was in contrast to A3G which caused inactivation of 84% of the clones and left only 3% of clones active ( Figure 4A ) . A3G also did not mutate some prot clones , but only 13% of the time ( Figure 4A ) . A3G did induce protease drug resistant mutations in 10% of the population ( Table S2 , D30N ) , but the examination of these clones individually demonstrated that they were inactivated by other mutations . Overall , we found that per mutation A3G was more likely to cause an inactivating mutation than A3F . This appeared to be due to the ability of a 5′CCC motif to cause more nonconservative mutations than 5′TTC in the prot ( Table S2 ) . For example , A3G had clear hot spots that caused inactivation of the protease , e . g . , W42 STOP , 20% of clones; G51R , 36% of clones; G52S , 52% of clones ( Table S2 ) . To investigate the A3F DNA scanning mechanism further we made mutants in A3G and A3F to alter their processive scanning behavior . For A3G , the only other A3 double Z-domain enzyme studied with regards to processivity , the NTD domain acts as a processivity factor [34] . The A3G CTD domain alone is non-processive ( Figure 5A–B and [53] , [54] ) . In order to focus in on the determinants of processivity , we recombinantly expressed the CTD domain of A3F and tested its processivity using ssDNA substrates as in Figure 1 . We found that the CTD of A3F could not processively deaminate cytosines that were spaced 63- or 30- nt apart , similar to the CTD of A3G ( Figure 5A–B , absence of 5′C & 3′C band ) . The A3F CTD could also not processively deaminate target cytosines 14- or 5-nt apart ( Figure 5C–D , absence of 5′C & 3′C band ) , similar to the full-length A3F enzyme ( Figure 1B–C ) . These data indicated that the NTD of A3F was a processivity factor . To determine the specific amino acids within the NTD that differentiate the processive scanning behaviors of A3F and A3G we aligned their amino acid sequences and looked for differences in the predicted helix 6 and loop 7 ( Figure S6 ) since these regions have been shown to influence the scanning behavior of A3G [34] . Specifically , it was found that helix 6 mediated sliding movements and loop 7 mediated jumping movements [34] . Since we could not observe any scanning by sliding for A3F ( Figure 1B–C ) , we hypothesized that residues within or near predicted N-terminal helix 6 would be different from A3G . For A3G , His186 was found to be essential for sliding movements [34] . Although A3F has a His181 equivalent to A3G ( His186 ) , at the end of the predicted helix 6 in the connection domain between the NTD and CTD , A3F has an additional three amino acids , 190NPM192 , in comparison to A3G ( Figure 6A ) . To test whether the 190NPM192 motif prevents A3F from sliding , we inserted the NPM motif into the equivalent position in A3G ( 195NPM197 ) creating an A3G NPM mutant . We then tested if A3G NPM was still able to undergo scanning by sliding . Using the ssDNA substrates with target cytosines close together enables the observation of processive deaminations by sliding [34] . On the substrate with cytosines separated by 5 nt , A3G NPM retained its processivity at an equivalent frequency to that of the wild-type A3G ( compare Figure 6B and Figure 1C , processivity factors ) . On the substrate with cytosines separated by 14 nt , A3G NPM was essentially not processive , as evidenced by a processivity factor of 1 which means that A3G NPM double deaminations occurred at the same frequency as expected if they were uncorrelated ( Figure 6C ) . This was in contrast to wild-type A3G that was able to processively deaminate cytosines located 14 nt apart ( Figure 1B , processivity factor 4 . 6 ) . These data indicated that the NPM insertion had decreased the sliding distance of A3G . To ensure that jumping was not affected , we tested the A3G NPM mutant on a substrate with cytosines separated by 63 nt without or with a complementary DNA or RNA annealed . First , we established the processivity on this substrate when fully single-stranded . Accordingly , the A3G NPM which had a decreased ability to slide , had a decreased processivity factor on this substrate in comparison to wild-type A3G ( compare Figure 6D , processivity factor of 5 . 1 to Figure 1D , processivity factor of 8 . 1 ) , but similar to A3F ( Figure 1D , processivity factor of 4 . 6 ) . When a complementary DNA was annealed the processivity of the A3G NPM was not decreased ( Figure 6E , processivity factor of 4 . 4 ) demonstrating that the jumping motion of A3G NPM was not affected . Similar results were found when a complementary RNA was annealed to the substrate ( Figure S2C ) . This was in contrast to the characteristic 2-fold decrease in processivity observed with A3G when a complementary DNA or RNA is annealed in between two target cytosines ( Figure 1D–E and Figure S2B ) consistent with the hypothesis that attempts to slide over the dsDNA region by wild-type A3G results in dissociation of the enzyme into the bulk solution . That we did not find an increase in the jumping efficiency for the A3G NPM ( Figure 6E ) , in contrast to A3F ( Figure 1E ) is in agreement with published data that suggest the determinants of jumping are separate from sliding and localized to predicted loop 7 ( Figure S6 and [34] ) . Further , the oligomerization state of A3G NPM is equivalent to wild-type A3G ( data not shown ) , not A3F and this may influence the jumping distance of an A3 enzyme ( Figure 2A–C ) . To ensure that the effects on processivity were due to specific changes to the residues interacting with ssDNA while scanning , rather than solely due to a poor affinity for ssDNA , we examined A3G NPM by circular dichroism ( CD ) spectroscopy and the ability of A3G NPM to bind ssDNA using rotational anisotropy . The CD analysis confirmed that A3G NPM and A3G were structurally similar ( data not shown ) . Interestingly , addition of the NPM residues to A3G resulted in a 2-fold increase in the binding affinity of A3G for the ssDNA , implicating these residues in the ssDNA-NTD interaction ( Table 1 , A3G , Kd of 130 nM; A3G NPM , Kd of 56 nM ) . The specific activity of A3G NPM was decreased ∼3-fold in comparison to A3G ( Table 4 , A3G , 15 pmol/µg/min; A3G NPM , 5 . 5 pmol/µg/min ) . To further investigate the influence of the NPM motif in A3F , we attempted a reciprocal mutation , i . e . , deleting the NPM motif from A3F . However , the mutant A3F did not express well in the Sf9 expression system indicating that the NPM deletion caused a structural instability . To circumvent this we made a conservative mutation in A3F to change the NPM motif to an NGM motif . We hypothesized that the Pro would have a significant influence on the functionality of the motif since Pro gives structural rigidity . We then tested the ability of the A3F NGM to processively deaminate two closely spaced deamination motifs by sliding . We found that A3F NGM was able to processively deaminate cytosines that were 5 nt and 14 nt apart ( Figure 7A–B , processivity factors of 2 . 1 and 2 . 4 ) , in contrast to A3F ( Figure 1B–C ) . When the distance between the cytosines was increased to 30 nt or 60 nt apart , A3F NGM was able to undergo processive deaminations similarly to A3F ( compare Figure 7C–D to Figure 1A and D ) . Interestingly , the apparent Kd of A3F NGM was 119 nM , which is 6-fold larger than the Kd of A3F ( Table 1 , 20 nM ) further implicating these residues in the enzyme-ssDNA interaction . The specific activity of A3F NGM was ∼1 . 5-fold higher than A3F ( Table 4 ) . The A3F NGM and A3G NPM results demonstrated that the presence of an NPM motif blocks the ability of both A3F and A3G to processively slide on ssDNA . Our model predicts that the A3G NPM mutant should be a poor inducer of mutagenesis during ( − ) DNA synthesis due the decreased ability of this mutant to slide on ssDNA ( Figure 6 ) . In agreement with the model , the A3G NPM induced mutagenesis poorly in the model HIV replication system ( Figure 8A ) , similar to A3F ( Figure 3B ) , but in contrast to wild-type A3G ( Figure 3A ) . The A3G NPM mutant had a mutation frequency in the HIV replication assay ( Table 2 , 0 . 29×10−2 mutations/bp ) , which was 9-fold less than wild-type A3G ( Table 2 , 2 . 63×10−2 mutations/bp ) . The spectrum and sequence analysis demonstrated that the sparse mutations induced by A3G NPM were still in 5′GG or 5′GGG contexts , but that much fewer occurred ( Figure 8A and Table S2 ) . The A3G NPM mutant rarely induced mutations in the prot ( Figure 8C ) and mutations in the lacZα region were less than A3F ( compare Figure 8D and Figure 3F ) . Notably , the A3G forms had a 100- ( A3G ) to 40- ( A3G NPM ) fold greater specific activity than A3F ( Table 4 ) . However , since A3G NPM and A3F similarly induced less mutations ( Figure 8A and Figure 3B ) than A3G ( Figure 3A ) , the data indicated that the ssDNA searching mechanism , but not the specific activity was a primary determining factor in levels of A3-induced mutagenesis . Our model , which is based on mutagenesis data from A3G , predicted that the mutation frequency of A3F NGM should increase in comparison to A3F . However , despite the A3F NGM mutant being able to slide ( Figure 7 ) , we found that A3F NGM remained inefficient at inducing mutagenesis in the in vitro HIV replication assay ( Figure 8B ) . The induced mutagenesis of A3F NGM ( Table 2 , 0 . 34×10−2 mutations/bp ) was more similar to A3F than A3G ( Table 2 ) . This could be due to A3F NGM sliding being ∼2-fold less efficient than A3G ( compare Figure 7A–B and Figure 1B–C ) or that the recovery of sliding alone in A3F is not sufficient for increasing the levels of mutagenesis . The latter possibility suggested another determining factor specific to A3F may affect its mutagenic ability . Namely , A3F NGM retained two distinct properties of A3F , the formation of tetramers and higher order oligomers ( data not shown ) and an increase in processivity with increasing distance between deamination motifs ( Figure 7 and data not shown ) . Therefore , we propose that the jumping mechanism of A3F that is retained in A3F NGM and distinct from that of A3G is detrimental to efficient mutagenesis and remains as such even the presence of sliding movements . With this being considered , the contributing factors to the efficiency of A3-induced mutagenesis is not only the balance between sliding and jumping , as exemplified by A3G , but also the type of jumping movements , as exemplified by A3F . The A3F NGM mutant also retained the 5′TTC specificity characteristic of A3F and induced similar mutations in HIV prot ( Table S2 ) . The biochemical data support the hypothesis that the processive scanning mechanism of the A3 enzyme can determine its mutagenic potential during reverse transcription . However , the in vitro HIV replication assay used in our experiments cannot account for how the HIV capsid environment may influence A3 enzyme-induced mutagenesis . Therefore , we used a single-cycle replication assay to test whether mutagenesis induced in the prot of HIVΔvif proviral DNA by the deamination activity of A3G , A3F and their mutant derivatives would recapitulate the results of A3-induced prot mutagenesis in the model HIV replication assay . In agreement with the biochemical data , in the HIV Δvif proviral DNA the A3G-induced mutations/kb were 6- to 8-fold higher than those of A3F , A3G NPM or A3F NGM ( Table 5 ) . Upon analysis of codon changing mutations , we found that the A3G hotspot in the prot was the Trp 42 codon , which was mutated to a stop codon in all clones containing a mutation , except one clone ( Table S3 ) . Clones mutated by A3G-catalyzed deaminations also contained other inactivating mutations such as G51R/E or G86R ( Table S3 ) . In regards to hotspots , the data were similar for A3G NPM , although fewer mutations were recovered ( Table 5 and Table S3 ) . These data supported our biochemical data in which the decrease in sliding by A3G NPM in comparison to A3G resulted in a decrease of mutagenic potential ( compare Figure 3A and Figure 8A ) . The prot clones exposed to A3F or A3F NGM had mutations that at best resulted in partial inhibition of protease activity , e . g . , D30N or M46I , and none that resulted in complete inactivation of protease activity ( Table S3 ) . It is interesting that A3F NGM induced ∼1 . 3-fold more mutations/kb than A3F , suggesting that there was a slight positive effect of the A3F NGM sliding ability on mutagenesis ( Table 5 ) . Overall , the prot sequencing data from HIV Δvif proviral clones was consistent with the conclusions from the in vitro model HIV replication assay and many deamination hotspots were common between the two assays ( compare Tables S2 and S3 ) . Differences may have resulted from different temporal dynamics of reverse transcription [59] and that the in vitro assay used a smaller segment of the prot gene . The observation from in vitro data that the 5′TTC motif was less able to cause inactivating mutations than the 5′CCC motif was consistent with HIV Δvif proviral DNA exposed to A3F or A3G ( Table S3 ) . Not only was the 5′CCC motif able to cause more inactivating mutations by overlapping with the Trp codon ( 5′TGG ) , which results in a stop codon , as previously observed [18] , but also because it was more likely to cause nonconservative mutations in comparison to the 5′TTC motif ( Table S3 and Ref [44] ) . The impact of A3G- and A3F-induced mutations on the infectivity of the proviral DNA was also examined using the eGFP reporter gene contained in the HIV pNL4-3 Δvif construct . Consistent with sequencing data from the prot region , the eGFP reporter gene of the integrated provirus from the same assays was inactivated 3- to 4-fold more in HIV Δvif virions exposed to A3G in comparison to A3F , A3G NPM or A3F NGM ( Figure 9A ) . To ensure this was not due to differences in encapsidation efficiency between these A3 enzymes we conducted quantitative immunoblotting on virions and cell lysates . Since we had transfected untagged A3 enzymes for these experiments to avoid the potential effects a tag may have on processivity ( Figure S7 ) , we initially standardized the antibodies for native A3G and A3F . Using equivalent amounts of purified protein and antibody dilutions , we determined that the antibody to A3F was 9-fold less sensitive than the antibody to A3G ( Figure 9B ) . As a result , we used this as a correction factor in the calculated amounts of these enzymes in virions and cells ( Figure 9C ) . The immunoblot results demonstrated that A3G and A3F were expressed in 293T cells and encapsidated into Δvif virions to a similar level ( Figure 9B–C ) . Therefore , the data support that there is a bone fide difference in the inherent mutagenic abilities of A3G and A3F . We also confirmed that A3G and its NPM mutant and A3F and its NGM mutant were expressed in cells and encapsidated in virions similarly ( Figure 9B–C ) enabling comparisons to be made between the mutant and wild-type forms of the enzymes . The analysis of A3G or A3F mutants from single-cycle infectivity assays was consistent with biochemical data . The A3G NPM mutant that had diminished sliding ability was less able to restrict HIV replication than A3G ( Figure 9A , 3-fold ) . A3F NGM was able to decrease HIVΔvif infectivity 10% more than A3F , suggesting a slight positive effect of its sliding ability , but this was not statistically significant ( Figure 9A ) . These data provided evidence that the processive scanning mechanism of the A3 enzyme influences the capacity to restrict HIV in a single cycle of replication . The disparity in HIV restriction efficiency was confirmed to be due to differences in mutational load by sequencing the HIVΔvif integrated provirus eGFP reporter gene ( Figure S8 ) .
Reports have demonstrated that A3F is less effective than A3G at restricting HIV replication and leaves less of a mutational footprint [22] , [23] , [24] , [29] . This could be due to many reasons such as differences in mRNA/protein expression levels [22] , [31] , virion encapsidation levels [4] , [33] , deamination site preference [4] , [18] , or the inherent biochemical characteristics of the enzymes that govern deamination activity during proviral DNA synthesis . There is no consensus in the literature regarding whether any of the variables determined by cellular conditions , e . g . , mRNA expression levels , create disparity between A3F and A3G HIV restriction activities . In addition , other reports have found an equal capacity of A3F and A3G to restrict HIV [3] , [5] , [6] , [7] , [20] , [21] , [30] . To account for these differences in the literature we undertook a biochemical characterization of A3F in comparison to A3G . The data have enabled us to form a biochemical model to account for cell-based observations and propose that the processive DNA scanning mechanism and the preferred deamination motif of A3 deoxycytidine deaminases are determinants of HIV restriction efficiency . The data support the hypothesis that a balanced sliding and jumping scanning mechanism is a major contributor to efficient restriction of HIV [34] and A3F has less potential to restrict HIV because it does not slide and uses a jumping translocation mechanism that is different than A3G ( Figure 1 and Figure 2 ) . Analysis of A3G and A3F mutants further support the model in which the mechanism that the enzymes scan DNA and not their specific activity can fully account for differences observed in HIV restriction ( Table 4 , Figure 3 , Figure 8 , and Figure 9 ) . In addition , A3F-induced mutations in preferred 5′TTC/5′TC motifs were less efficient at inducing gene inactivation than the preferred A3G 5′CCC/5′CC motifs , similar to what was identified for A3A ( prefers 5′TTC/5′TC ) [44] , adding another distinction in the mutagenic ability of A3F ( Tables S2 and S3 ) . However , the data cannot support that A3F has no effect on HIV since it is suppressed by Vif [6] , but there is evidence that the restriction abilities are distinct from A3G in regards to mutagenic load , selection pressure on HIV and contribution of deamination-independent HIV restriction [24] , [29] , [60] , [61] , [62] , [63] . It was initially recognized by Zennou and Bieniasz that per mutation , A3G could cause a much larger decrease in HIV infectivity than A3F [4] . This early study on A3F [4] was in contrast to other early studies published showing A3F was similar in effectiveness to A3G [3] , [5] , [6] , [7] , [30] . Such incongruent data still remains in the literature [21] , [22] , [23] , [24] , [29] and may be due to different experimental systems . Specific to our data , we observed that the HIVΔvif retained 51% infectivity in the presence of A3F and 13% infectivity in the presence of A3G , suggesting that A3F is not as effective as A3G at restricting HIV ( Figure 9A ) . However , Albin et al . found that over multiple replication cycles , A3F restricted HIV replication similarly to A3G and selected for Vif mutant revertants [61] . It may be that A3G requires only one exposure to HIV for high level restriction compared to A3F that may require multiple cycles for strong HIV restriction , but the end point is the same . Importantly , multiple infection rounds more closely mimics how A3 enzymes would interact with HIV in vivo . Nonetheless , our data propose that the mechanism by which A3G and A3F reach this end is different and that A3F has the potential to cause more sequence diversification of HIV than A3G . This idea is supported by Chaipan et al . that found that A3F suppressed HIV in multiple rounds of replication but required a longer period of exposure to HIV before the level of suppression reached that of A3G [24] . This is consistent with our sequence data from the prot of integrated proviruses ( Table S3 ) . As such , the role of A3F may be to supplement mutagenesis induced by A3G [21] , [29] since their effects have been shown to be additive [5] or be distinct from A3G and perhaps rely on a deamination-independent mechanism , such as inhibition of reverse transcription and integration [60] , [63] , [64] . A3F has been reported to exert a larger deamination-independent inhibition of HIV replication than A3G , but this is not as effective as deamination-mediated restriction of HIV [60] , [63] . To characterize the mechanism by which A3 enzymes induce mutagenesis we studied the A3G NPM mutant . The A3G NPM mutant demonstrated that the scanning mechanism on DNA and not specific activity is a primary determinant in mutation induction during reverse transcription . Since A3F had a lower specific activity than A3G ( Table 4 ) , it could be argued that this was contributing to the lower level of induced mutagenesis ( Figure 3 ) . However , the A3G NPM mutant , which had decreased sliding in comparison to wild-type A3G ( compare Figure 1 and Figure 6 ) , retained a specific activity more similar to that of A3G than A3F ( Table 4 ) , but induced a very low level of mutagenesis ( Figure 8A , C–D ) and decreased HIV infectivity only 2-fold versus A3G that decreased HIV infectivity 8-fold ( Figure 9A ) . These data suggest that specific activity is not a determinant in the ability to cause mutations during reverse transcription and is supported by previous data in which the specific activity of the enzyme was inconsequential during reverse transcription [43] . This appears to be because the activity of the enzyme during reverse transcription is instead determined by factors such as ( − ) DNA synthesis and RNaseH activity [43] . The A3G NPM mutant further confirmed that the determinants within the NTD for sliding involve residues near A3G predicted helix 6 and that this is distinct from the determinant for jumping ( Figure 6D–E ) , which in A3G is loop 7 ( Figure S6 ) . Of note , the NPM motif is predicted to be at the end of NTD helix 6 , which is a connection point between the NTD and the CTD domains [65] , [66] . Through amino acid sequence alignment we identified that A3D is the only other double domain A3 deaminase to contain an NPM motif at the end of predicted helix 6 , suggesting that A3D would also lack sliding movements while scanning ssDNA , similarly to A3F . Some specific residues within helix 6 have previously been shown to affect specific activity [34] , [66] , possibly because of structural changes in the connection between the NTD and CTD that can affect the catalytic activity of the CTD or DNA binding affinity . Insertion of the NPM motif into A3G immediately after the predicted helix 6 ends ( Figure 6A ) did not cause a large disruption in structure based on CD spectra ( data not shown ) , but did result in a ∼3-fold decrease in specific activity ( Table 4 ) and ∼2-fold increase in binding affinity for ssDNA ( Table 1 ) . To confirm a role of the NPM motif in blocking sliding movements we mutated this region in A3F to create an A3F NGM mutant with the hypothesis that removing the rigid proline residue would enable the enzyme to slide on ssDNA and deaminate closely spaced residues . Consistent with the hypothesis , closely spaced residues were processively deaminated by A3F NGM ( Figure 7A–B ) . However , the ability to slide did not enable A3F NGM to induce high levels of mutagenesis similar to A3G in vitro ( Figure 8 ) or in a single-cycle infectivity assay ( Figure 9A and Table 5 ) . This does not preclude that jumping and sliding are important for inducing mutagenesis in virus infected cells since the A3G NPM mutant that had decreased sliding restricted HIV similarly to A3F in single cycle infectivity assays ( Figure 9A ) . Rather , these data indicated that the ability to slide and jump is necessary , but not sufficient to induce high levels of mutagenesis . The data supported the conclusion that the type of sliding and jumping movements , e . g . , distance transversed was also important . Namely , we found that A3F processivity on ssDNA increased with increasing distance between deamination motifs , in contrast to A3G , demonstrating that the average jumping distance of A3F was larger than A3G ( Figure 2A–B ) . This was confirmed with sequence analysis from the model HIV replication assay in which a larger number of deaminations were >20 nt apart for A3F than A3G ( Table 3 ) . Thus , the A3F NGM mutant could slide , but was not truly a mimic of A3G . All together , it appears that the sliding and jumping mechanism of A3G is specifically optimal to induce a large number of deaminations during reverse transcription of DNA . An important note regarding the study of A3F is that we found N-terminally tagged GST-A3F was not processive ( Figure S7 ) , despite binding ssDNA with a Kd of 46±4 nM . That the binding affinity of GST-A3F was more similar to A3F ( Table 1 , Kd of 20 nM ) than A3F CTD ( Table 1 , Kd of 288 nM ) indicated that the GST-A3F was able to bind ssDNA with both NTD and CTD domains , despite a lack of processivity . This suggested that the GST tag caused steric hindrance on amino acid determinants for processivity in the NTD . Interestingly , we observed that nonprocessive A3F forms , both A3F CTD and GST-A3F , induced more mutations than wild-type , processive A3F ( compare Figure 3 and Figures S9 and S10 ) . We also found that A3A , which is largely nonprocessive , induced slightly less mutations than A3G in the in vitro HIV replication assay [44] , but more than A3F . Although this initially seems difficult to reconcile , it is consistent with the overall hypothesis that processivity is related to mutagenic potential , since processive A3G is still the most efficient at inducing mutagenesis . It is only that a lack of processivity appears to be better than an “ineffective” processive enzyme such as A3F . This is not due to differences in the assay systems for characterizing processive deaminations on ssDNA oligonucleotides and the model HIV replication assay since addition of NC and RT to the ssDNA oligonucleotides in a deamination reaction did not change our observations regarding A3F CTD processivity ( Figure S11 ) . In comparison to the nonprocessive A3F CTD and GST-A3F , processive A3F leaves many potential deamination motifs unmodified ( compare Figure 3 and Figures S9 and S10 ) . Although there is inefficiency in the GST-A3F and A3F CTD having to dissociate and reassociate with the substrate many times , the reassociations can be much closer to the previous dissociation resulting in a more thorough search of the DNA . For example , we found that in the model HIV replication assay , 61% of A3F CTD-induced mutations were >20 nt apart in contrast to A3F where 75% of induced mutations were >20 nt apart ( Table 3 and data not shown ) . Since the HIV replication assay is not conducted under single hit conditions , the results emphasize the inefficiency of the searching mechanism used by A3F . Since the binding affinity of A3F for ssDNA is tighter than A3G or A3F CTD ( Table 1 ) , it is conceivable that A3F may also have a lack of frequent movements or excursions on the ssDNA that contribute to the inefficient search for deamination motifs . However , resolution of this speculation awaits single-molecule analysis . In sum , the data demonstrated that the interactions of A3F with ssDNA are essentially detrimental to its ability to induce a high mutation frequency . Our data demonstrate two main points . First , the data provide a biochemical reason for the inefficiency with which A3F-induces mutagenesis of HIVΔvif as observed in this report and by others [22] , [23] , [24] by demonstrating that the processive scanning behavior of A3F is detrimental to its mutagenic potential . The data establish that a balanced sliding and jumping ssDNA scanning mechanism similar to A3G is required for the most efficient induction of HIV mutagenesis . Secondly , the data show that deamination of 5′CCC/5′CC has more gene inactivating potential than 5′TTC/5′TC providing an additional reason for less restriction of HIV by A3F than A3G , in agreement with previous reports [4] , [18] , [44] . The data does not preclude that A3F can effectively restrict HIV and is in agreement with studies showing that A3F can restrict HIV in multiple rounds of infection [24] , [61] , but since the number of mutations induced has been correlated to HIV inactivation [45] , [67] , the data support the interpretation that A3F inactivates HIV less efficiently than A3G in a single round of infection .
Recombinant baculovirus production for expression of GST-A3G , GST-A3F ( NCBI Accession BC038808 ) , GST-A3G CTD ( amino acids 197–380 ) , GST-A3F CTD ( amino acids 195–373 ) , GST-A3G NPM , GST-A3F NGM or GST-nucleocapsid protein ( NC ) in Sf9 cells was carried out using the transfer vector pAcG2T ( BD Biosciences ) , as previously described [34] , [35] , [54] . Site directed mutagenesis was used to create the A3G NPM and A3F NGM clones . Cloning primers for A3 enzymes and the site directed mutagenesis primers were obtained from Integrated DNA Technologies and are listed in Table S1 . Sf9 cells were infected with recombinant virus at a multiplicity of infection ( MOI ) of 1 , except for GST-A3F and GST-A3F CTD which were infected at an MOI of 2 . Recombinant baculovirus infected Sf9 cells were harvested after 72 h of infection . Cells were lysed in the presence of RNaseA and the proteins ( A3G , A3G NPM , A3G CTD , and NC ) were purified as described previously [54] to obtain protein that was cleaved from the GST tag and 95% pure . The A3F , A3F NGM , and A3F CTD enzymes were eluted from the glutathione-sepharose resin ( GE Healthcare ) with the GST tag , as previously described [35] . The samples were then treated with thrombin ( Merck Millipore; A3F and A3F NGM , 0 . 02 U/µL; A3F CTD , 0 . 10 U/µL ) for 2–5 hours at 21°C to cleave the GST tag . A DEAE Fast Flow column ( GE Healthcare ) was then used to purify the A3F , A3F NGM , and A3F CTD from the GST tag and thrombin . The proteins were loaded in low salt buffer containing 50 mM Tris pH 8 . 0 , 50 mM NaCl , 10% glycerol , and 1 mM DTT . A linear gradient from 50 mM NaCl to 1 M NaCl was used to differentially elute the enzymes . The enzymes eluted at approximately 450 mM NaCl and were 90% pure . The SDS-PAGE gels of the purified A3 enzymes are shown in Figure S12 . Protein fractions were stored at −80°C . HIV RT ( p66/p51 ) [68] was generously provided by Dr . Stuart F . J . Le Grice ( NCI , National Institutes of Health ) . The oligomerization state of A3 enzymes was determined by subjecting 10–15 µg of the purified enzymes to size exclusion chromatography using a 10 mL Superdex 200 ( GE Healthcare ) resin bed contained in a column with a 0 . 5 cm diameter and 16 cm height . The running buffer used was 50 mM Tris pH 7 . 5 and 200 mM NaCl . The Bio-Rad gel filtration standard set was used to generate a standard curve from which molecular masses and oligomerization states were calculated . A3-induced mutagenesis of ssDNA during reverse transcription of an RNA template was measured using an in vitro assay , which models reverse transcription from an RNA template and second strand synthesis , and was performed as described previously [34] . Briefly , a synthetic ( + ) RNA is synthesized that contains a polypurine tract ( PPT ) , 120 nt of the catalytic domain of the HIV protease ( prot ) , and lacZα ( 248 nt ) . The PPT is used as a primer for ( + ) DNA synthesis and enables synthesis of dsDNA . The lacZα serves as a reporter gene for mutations by blue/white screening . The HIV protease gene was obtained by PCR using clone p93TH253 . 3 obtained through the AIDS Research and Reference Reagent Program , Division of AIDS , NIAID , NIH from Dr . Feng Gao and Dr . Beatrice Hahn [69] . The RNA template ( 50 nM ) was annealed to a 24 nt DNA primer [34] and incubated with NC ( 1 . 5 µM ) , RT ( 1 . 2 µM ) , and dNTPs ( 500 µM ) in RT buffer ( 50 mM Tris pH 7 . 4 , 40 mM KCl , 10 mM MgCl2 , 1 mM DTT ) in the presence or absence of 200 nM of A3G , A3F , A3G NPM or A3F NGM . Synthesized dsDNA was PCR amplified using Pfu Cx Turbo Hotstart ( Agilent Technologies ) that can use uracils as a template with high fidelity . The amplicons were cloned into a pET-Blue vector backbone that would allow the experimentally synthesized lacZα to be used for α-complementation [34] . At least twenty-five mutated clones for each condition tested were analyzed . DNA sequencing was carried out at the National Research Council of Canada ( Saskatoon , Saskatchewan ) . A t-test was used for statistical analysis of sequences . The ssDNA substrates were obtained from Tri-Link Biotechnologies and are listed in Table S1 . Deaminations were detected by resolving Fluorescein ( F ) -labeled DNA that had been treated with Uracil DNA Glycosylase ( New England Biolabs ) and heated under alkaline conditions on a 10% , 16% , or 20% v/v denaturing polyacrylamide gel , as described previously [35] . The gel type was determined by fragment sizes produced by each substrate . Reactions were carried out under single hit conditions , i . e . , <15% substrate usage [46] , to ensure that a single ssDNA substrate was interacting with at most a single enzyme . Under these conditions , a processivity factor can be determined by comparing the total number of deaminations occurring at two sites on the same DNA substrate to a calculated theoretical value of the expected deaminations that would occur at those two sites if the deaminations were not processive ( see reference [35] ) . In order to obtain substrate usage within this range under steady-state conditions , the enzyme and DNA concentration were varied based on the enzyme specific activity . More ssDNA was used with A3G to ensure clear observation of all deamination bands despite the large preference for the 5′C . However , the data are not altered with ssDNA concentration ( data not shown ) . For A3G and A3G NPM , 30 , 40 , or 100 nM enzyme was incubated with 300 or 500 nM fluorescein ( F ) -labeled ssDNA . For A3F , A3F NGM , and A3F CTD , 100 nM enzyme was incubated with 50 or 100 nM F-labeled ssDNA . For A3G CTD , 1000 nM enzyme was incubated with 500 nM F-labeled ssDNA . Reactions were incubated at 37°C for 1–50 min . Gel pictures were obtained using a Typhoon Trio ( GE Healthcare ) multipurpose scanner and analysis of integrated gel band intensities used ImageQuant software ( GE Healthcare ) . The specific activity was calculated from single-hit condition reactions by determining the picomoles of substrate used per minute for a microgram of enzyme . Steady state fluorescence depolarization ( rotational anisotropy ) was used to measure enzyme-ssDNA binding affinities using the same F-labeled ssDNA substrates ( with cytosines 63 nt apart ) that were used for deamination reactions ( Table S1 ) . Reactions were 60 µL and contained F-labeled ssDNA ( 10 nM ) in RT buffer and A3G ( 0–650 nM ) , A3F ( 0–80 nM ) , A3F CTD ( 0–600 nM ) , A3G NPM ( 0–350 nM ) , or A3F NGM ( 0–650 nM ) were titrated into the reaction . A QuantaMaster QM-4 spectrofluorometer ( Photon Technology International ) with a dual emission channel was used to collect data and calculate anisotropy . Measurements were made at 21°C . Samples were excited with vertically polarized light at 495 nm ( 6 nm band pass ) and vertical and horizontal emissions were measured at 520 nm ( 6 nm band pass ) . Apparent dissociation constants ( Kd ) were obtained by fitting to a sigmoidal curve using Sigma Plot 11 . 2 software . VSV-G pseudotyped HIV pNL4-3 Δvif viruses were produced by transfecting 3×105 293T cells per well in a 6-well plate with Qiagen Polyfect reagent . Specifically , transfections used 1100 ng of pHIVΔvif [70] , which expresses an eGFP reporter gene and 630 ng of pLTR-G ( Addgene ) , which expresses the VSV-G protein , in the presence or absence of 220 ng of A3G , A3F or A3F NGM or 350 ng of A3G NPM in pcDNA3 . 1 . The transfections used empty pcDNA3 . 1 to achieve equivalent amounts of DNA . The cotransfection molar ratio of A3 enzymes in pcDNA3 . 1 to the pNL4-3 Δvif was 0 . 33:1 ( A3G , A3F , or A3F NGM ) or 0 . 59:1 ( A3G NPM ) . The A3G ( cat# 9952 ) and A3F ( cat # 10100 ) expression plasmids were obtained from the NIH AIDS Reagent program with C-terminal tags . A stop codon was introduced immediately after the A3G or A3F coding sequence to enable expression of native A3 enzymes . The amino acid sequence of the A3G and A3F clones were identical to those used in biochemical assays . Subsequently , site directed mutagenesis was used to create the A3G NPM and A3F NGM clones . The site directed mutagenesis primers were obtained from Integrated DNA Technologies and are listed in Table S1 . Sixteen hours after the transfection , the cells were washed with PBS and the medium replaced . Virus-containing supernatants were collected 48 hours after the media change and filtered through 0 . 22 µm syringe filters . Virus was quantified by a p24 enzyme-linked immunosorbent assay ( QuickTiter Lentivirus Titer Kit , Cell Biolabs Inc . ) . Target 293T cells were infected at an MOI of 0 . 5 by spinoculation at 800× g for 1 h in the presence of 8 µg/ml of polybrene [71] . Infection levels in 293T cells was determined by flow cytometry by detecting eGFP fluorescence at 48 hours post infection and data were normalized to HIVΔvif infections in the absence of A3 enzymes . Infected 293T cells were harvested after 48 h and the DNA was extracted using the Qiagen DNeasy Blood and Tissue kit . DNA was treated with DpnI ( New England Biolabs ) to remove possible contaminating plasmid DNA and the prot ( nt 2280-2631 ) sequences were amplified by PCR using Phusion High Fidelity Polymerase ( New England Biolabs ) . Primers were obtained from Integrated DNA and are listed in Table S1 . PCR products were purified and cloned with the Zero Blunt TOPO PCR cloning kit ( Invitrogen ) . DNA sequencing was carried out at the National Research Council of Canada ( Saskatoon , Saskatchewan ) . The A3G and A3F enzymes were detected in cell lysates ( 40 µg total protein ) and virions ( 130 ng of p24 ) used for single-cycle infectivity assays using antibodies to the native enzymes . For A3G we used the ApoC17 rabbit antiserum ( Cat # 10082 , NIH AIDS Reagent Program ) and for A3F we used the C-18 polyclonal rabbit antibody ( Cat # 11474 , NIH AIDS Reagent Program ) . Loading controls for cell lysates ( α-tubulin , Sigma ) and virions ( p24 , Cat #3537 , NIH AIDS Reagent Program ) were detected using mouse monoclonal antibodies . Proteins of interest and loading controls were detected in parallel on the same gel by using the Licor/Odyssey system ( IRDye 680-labeled goat anti-rabbit secondary antibody and IRDye 800-labeled goat anti-mouse secondary antibody ) . Visualization with an Odyssey Infrared Imaging System ( Licor ) and analysis of bands with Odyssey software enabled intensities of bands to be determined . Analysis of a titration of purified A3G and A3F with their respective antibodies showed that A3F detection was 9-fold less sensitive than A3G detection at a 1/1000 antibody dilution . Further , doubling the amount of antibody to A3F ( 1/500 ) resulted in a doubling of the A3F detection sensitivity in comparison to the antibody to A3G ( 1/1000 ) . Therefore , an appropriate correction factor for the antibody dilution was used to adjust the integrated band intensities of A3F to enable comparison with A3G . Antibodies were used at a dilution of 1/1000 except for A3F or A3F NGM containing cell lysates which required a dilution of 1/500 for detection of A3F or A3F NGM . A t-test was used for statistical analysis . | Human cells possess a family of seven DNA-modification enzymes , termed APOBEC3 , that function as part of our innate immune system . The enzymes modify cytosine in DNA which induces mutations . There are particular enzymes , APOBEC3D , APOBEC3F , APOBEC3G and APOBEC3H , that appear to be most relevant to restricting HIV-1 replication in CD4+ T cells using this mutagenic mechanism , if they can avoid degradation that is induced by the HIV-1 protein Vif . There has been little biochemical analysis of APOBEC3 enzymes other than APOBEC3G in terms of the mechanism by which these enzymes search DNA for target cytosines to deaminate . We conducted a biochemical analysis of APOBEC3F . We found that while APOBEC3G uses 1-dimensional sliding and 3-dimensional translocations , APOBEC3F is restricted to 3-dimensional translocations . This makes the searching mechanism of APOBEC3F superficial and detrimental to the induction of a large number of mutations . In addition , gene inactivation was less likely to occur upon deamination of the target motif of APOBEC3F ( 5′TTC ) , in comparison to the target motif of APOBEC3G ( 5′CCC ) . All together the data support a model in which the way these enzymes scan DNA can predict the magnitude of mutagenesis induced and the target motif can predict ability to cause gene inactivation . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"immunodeficiency",
"viruses",
"biochemistry",
"dna",
"modification",
"nucleic",
"acids",
"virology",
"enzymes",
"dna",
"biology",
"enzyme",
"kinetics",
"microbiology",
"viral",
"replication",
"viral",
"nucleic",
"acid"
] | 2014 | Different Mutagenic Potential of HIV-1 Restriction Factors APOBEC3G and APOBEC3F Is Determined by Distinct Single-Stranded DNA Scanning Mechanisms |
Mounting evidence suggests the interaction between stress and genetics contribute to the development of depressive symptoms . Currently , the molecular mechanisms mediating this process are poorly understood , hindering the development of new clinical interventions . Here , we investigate the interaction between neuropsin , a serine protease , and chronic stress on the development of depressive-like behaviours in mice . We found no difference in baseline behaviour between neuropsin knockout and wild-type mice . However , our results show that neuropsin knockout mice are protected against the development of depressive-like behaviours and memory impairment following chronic stress . We hypothesised that this difference in behaviour may be due to an interaction between neuropsin and elevated plasma corticosterone . To test this , we subjected mice to chronic corticosterone injections . These injections resulted in changes to hippocampal structure similar to that observed following chronic stress . We found that inactivation of neuropsin limits the extent of these anatomical changes in both the chronic stress and the corticosterone injection exposed cohorts . We next used viral vectors to knockdown or overexpress neuropsin in the hippocampus to confirm the results of the KO study . Additionally , we found that inactivation of neuropsin limited glutamate dysregulation , associated with increased generation of reactive oxygen species , resulting from prolonged elevated plasma corticosterone . In this study , we demonstrate that neuropsin inactivation protects against the impairment of hippocampal functions and the depressive-like behaviour induced by chronic stress or high levels of corticosterone . Consequently , we suggest neuropsin is a potential target for clinical interventions for the management of stress disorders .
Stress , especially chronic stress , can lead to disorders such as depression and consequently result in significant morbidity and reduced quality of life [1] . Despite considerable research , the mechanisms underlying stress-related illnesses are unclear . Neuropsin ( NP , also named Klk8 ) is a serine protease that exhibits trypsin-like activity with a strong preference for Arg in the P1 position [2] . Its substrates include neuregulin-1 , fibronectin , vitronectin , synaptic adhesion molecule L1 , and Ephrin type-B receptor 2 ( EphB2 ) [3–6] . Neuropsin is predominantly expressed in the amygdala and CA regions of the hippocampus [6–8] . These two regions have been long associated with stress , emotion , learning and memory processes [9–11] . There are several pieces of evidence suggesting a link between neuropsin and stress . First , stress increases the expression of neuropsin mRNA in the hippocampus through a glucocorticoid-sensitive pathway [12] . Second , neuropsin has been shown to affect stress-induced anxiety by cleaving the extracellular portion of EphB2 [6] . Finally , the antidepressant fluoxetine alters the expression of neuropsin in the hippocampus [13] . However , the importance of neuropsin in chronic stress induced impairment of brain function and behaviour still remains unknown . In this study , we sought to investigate the role of neuropsin in the regulation of depressive-like behaviour , learning and memory following chronic stress . It is well known that stress induces a rapid rise in corticosterone . Chronically elevated corticosterone impairs memory , reduces neuronal spine density , and decreases hippocampal neurogenesis and volume [14–17] . Therefore , we further explored the interaction between corticosterone and neuropsin and the associated effects on behaviour , gene expression and hippocampal architecture . We investigated the role of neuropsin using multiple models: a knock out; a lenti-viral vector induced knock down; and an AAV vector induced overexpression model . Our results suggest that understanding the role of neuropsin in the brain may open new possibilities for the treatment of stress-associated disorders .
To investigate whether neuropsin expression is influenced by stress , we examined the expression of neuropsin in the hippocampus following acute and chronic restraint exposure . Eight week old C57BL/6 male mice were subjected to restraint stress for 30 min to model acute stress ( n = 4 ) or for 28 days to model chronic stress ( n = 8 ) . Mice were killed and hippocampal neuropsin mRNA was quantified with real-time PCR . The data show hippocampal neuropsin expression was increased after both acute restraint stress ( mean ± SEM; control: 1 . 00 ± 0 . 09; stress: 1 . 5 ± 0 . 17 , p = 0 . 036 , t = 2 . 68 ) and 28 days chronic stress ( control: 1 . 00 ± 0 . 192; stress: 1 . 94 ± 0 . 24 , p = 0 . 029 , t = 2 . 53 ) . Next , we used neuropsin knockout ( KO ) mice ( n = 8 ) and wild-type ( WT ) mice ( n = 10 ) to conduct tests to assess depressive-like behaviour; this included novelty suppressed feeding , forced swim , and a sucrose preference test . Our data showed there was no difference between WT and KO mice for depressive-like behaviour ( Fig 1A , 1C and 1E ) . Whilst a previous study suggested that neuropsin KO mice exhibit less anxiety compared to WT following acute restraint stress [6] , the role of neurospin in chronic stress remained unclear . To study this , both KO ( n = 9 ) and WT ( n = 10 ) mice were subjected to a battery of restraint stress , footshock and sleep deprivation for 28 days ( one stress paradigm per day ) . Behavioural tests were started on the 12th day of the 28 day stress protocol . We first measured the activity of mice using an open field test . The result showed there was an effect from stress but there was no significant difference between the WT and KO groups ( effect of stress , F1 , 34 = 10 . 33 , p = 0 . 003; NP , p = 0 . 28 ) ( S1A Fig ) . In the novelty suppressed feeding test , KO mice approached food pellets quicker than WT mice ( p = 0 . 001 , t = 3 . 83 ) . For the forced swim test , the immobility time of KO mice was less than WT ( p = 0 . 031 , t = 2 . 37 ) . Additionally , neuropsin KO mice drank more sucrose in the sucrose preference test ( p = 0 . 0054 , t = 3 . 18 ) ( Fig 1A , 1C and 1E ) . These results indicate that neuropsin KO mice showed less depressive-like behaviour following chronic stress compared to littermate WT mice . Memory is affected by numerous factors including stress and depression . Neuropsin has been suggested to be essential for memory acquisition , although some inconsistencies exist in the literature [18–21] . To further investigate whether neuropsin is involved in learning and memory following chronic stress , we used two independent cohorts of mice to perform a water maze test . First , we measured the basal level of learning and memory . WT and KO mice ( n = 11 ) had four trials a day for three days . Assessments of spatial memory were obtained by placing mice in the water maze without a platform two weeks after training and recording the time spent in the target zone and the number of times they crossed the phantom platform location . Results show there was no difference between the two training curves ( escape latency ) ( Repeated measures ANOVA , p = 0 . 69 , F11 , 220 = 0 . 16 ) . In the memory test , there were also no differences in the time spent in the target zone ( p = 0 . 33 , t = 0 . 98 ) and the number of phantom platform crosses ( p = 0 . 68 , t = 0 . 41 ) ( Fig 2A and 2B ) . These data indicate that neuropsin inactivation does not affect basal learning and memory as measured in a water maze . Next , we assessed spatial learning and memory in another cohort of mice ( n = 10 ) which had experienced chronic stress ( training started from day 12 ) . Results showed no difference in the training curves ( p = 0 . 81 , F11 , 187 = 0 . 054 ) . However , the memory test showed KO mice spent more time in the target zone ( p = 0 . 0018 , t = 3 . 69 ) and made more phantom platform crosses ( p = 0 . 0069 , t = 3 . 07 ) ( Fig 2A and 2B ) . These results suggest that in mice exposed to chronic stress , a lack of neuropsin does not influence spatial learning , but does improve indicators of memory . Neurogenesis has the effect of buffering stress and depressive-like behaviours and has long been associated with learning and memory [22 , 23] . Previous research has demonstrated the status of the CA3 region can affect the survival of newly born neurons in the dentate gyrus [24] . Here , we asked whether neuropsin plays a role in the regulation of adult hippocampal neurogenesis following chronic stress . Mice were divided into four groups ( WT , KO , n = 7; WT , KO with chronic stress , n = 10 in each group ) . The day prior to the commencement of the chronic stress protocol , mice were injected with 5-bromo-2'-deoxyuridine ( BrdU ) ( 200 mg/kg , i . p . ) to label newly born cells . Mice were killed 28 days after BrdU injection to evaluate cell survival of new neurons generated at the time of BrdU injection . To assess proliferation of neuronal progenitors at the time of culling , we stained for doublecortin ( DCX ) , a marker for immature neurons , and KI67 , a marker for cell proliferation . A two-way ANOVA revealed a significant stress effect on neurogenesis ( DCX: F1 , 29 = 30 . 7 , p < 0 . 0001; BrdU: F1 , 29 = 389 . 7 , p < 0 . 0001 , KI67: F1 , 29 = 7 . 6 , p = 0 . 01 ) . For the DCX , BrdU , and KI67 cell counts in the dentate gyrus , there were no significant differences between KO and WT mice ( p > 0 . 05 ) . In the chronic stress groups , we detected more DCX ( p = 0 . 002 , t = 3 . 3 ) and BrdU ( p < 0 . 0001 , t = 4 . 71 ) positive cells in the dentate gyrus of KO mice compared to WT mice . However , there was no significant difference in KI67 cell count ( p = 0 . 86 , t = 0 . 17 ) ( Fig 3A–3C ) . Overall , these results suggest neuropsin does not affect neurogenesis under normal conditions but does play a role , especially in cell survival , when mice experience chronic stress . Corticosterone is released in response to stress . To confirm whether the stress models we used increased plasma corticosterone , we measured basal levels in both neuropsin WT and KO mice ( n = 8 ) and levels after 3 hrs of restraint stress , 30 min after footshock , after 6 hrs of sleep deprivation and 24 hrs after all stressors were completed . Our result showed that the stressors we used in this study elevated plasma corticosterone ( Two way ANOVA; F4 , 56 = 91 . 24 , p < 0 . 0001 ) and there was no significant difference between WT and KO mice ( basal , p = 0 . 69 , t = 0 . 4; restraint , p = 0 . 15 , t = 1 . 5; footshock , p = 0 . 77 , t = 0 . 29; sleep deprivation , p = 0 . 16 , t = 0 . 48; 24 hrs after stress , p = 0 . 34 , t = 0 . 98 ) ( S2 Fig ) . In our experiment , we sought to establish whether our findings on behaviour after chronic stress were due to an interaction between elevated corticosterone and neuropsin inactivation . To do so , KO and WT mice were injected with oil ( KO/oil , n = 12 and WT/oil , n = 10 ) or corticosterone ( KO/CORT , n = 8 and WT/CORT , n = 10 , 20 mg/kg/day , subcutaneously ) daily for 32 days to imitate the elevated corticosterone during chronic stress . On day 12 , we conducted a battery of tests for assessing depressive-like behaviour , learning and memory . No differences were found in any of the behavioural tests from the oil injected groups . ( Figs 1B , 1D , 1F and 2C and 2D ) . However , in the corticosterone injected groups , we found that KO mice exhibited less depressive-like behaviour than WT mice . Corticosterone significantly decreased activity but there was no significant difference between WT and KO groups ( effect of corticosterone , F1 , 37 = 0 . 3 , p < 0 . 0001; effect of neuropsin P = 0 . 58 ) ( S1B Fig ) . KO mice had a reduced latency to approach food pellets in the novelty suppressed feeding test ( p = 0 . 002 , t = 3 . 54 ) , reduced immobility time in the forced swim test ( p = 0 . 039 , t = 2 . 22 ) and drank more sucrose than WT mice in the sucrose preference test ( p = 0 . 04 , t = 2 . 18 ) ( Fig 1B , 1D and 1F ) . On day 17 , mice were trained in the water maze , and memory performance was assessed two weeks later . There was no difference in the training curves between WT and KO ( p = 0 . 54 , F11 , 176 = 0 . 39 ) . However , for the memory test , the KO mice spent more time in the target zone ( p = 0 . 026 , t = 2 . 44 ) and made more phantom platform crosses ( p = 0 . 006 , t = 3 . 11 ) ( Fig 2C and 2D ) . These results suggest that corticosterone interacts with neuropsin to affect spatial memory performance . Evidence shows that hippocampal neurogenesis is inhibited by corticosterone [25] . Consequently , we sought to ascertain whether neuropsin interacts with corticosterone to regulate neurogenesis . Our data shows that corticosterone injection strongly reduced hippocampal neurogenesis ( DCX: F1 , 20 = 131 . 6 , p < 0 . 0001; BrdU: F1 , 20 = 71 . 96 , p < 0 . 0001; KI67: F1 , 20 = 75 . 64 , p < 0 . 0001 ) . However , there was a significantly higher level of DCX ( p = 0 . 043 , t = 2 . 15 ) and BrdU ( p = 0 . 003 , t = 3 . 38 ) cell counts , but not KI67 ( p = 0 . 19 , t = 1 . 34 ) in the dentate gyrus in the neuropsin KO compared to WT ( Fig 3A–3C ) . This suggests neuropsin interacts with plasma corticosterone to influence cell survival but not proliferation in the dentate gyrus . In regards to dendritic morphology , it is known that elevated corticosterone reduces hippocampal dendritic spine density [26] , which has been suggested to be important for learning and memory [27] . To better understand whether neuropsin interacts with corticosterone to affect dendritic structure , we utilised Golgi staining to analyse the spine density in CA3 . Our data showed corticosterone reduced spine density significantly ( F1 , 12 = 127 . 3 , p < 0 . 0001 ) . There was no significant difference in spine density between WT and KO mice with oil injection ( p = 0 . 42 , t = 0 . 82 ) . In the corticosterone injected groups , KO mice exhibited a higher spine density than WT mice ( p = 0 . 0002 , t = 5 . 22 ) ( Fig 4A and 4B ) . Stress may accelerate the ageing process and accelerate neurodegeneration [28] . Previous studies have shown that neuropsin is expressed in oligodendrocytes after injury to the central nervous system [29] and that prolonged corticosterone treatment of adult rats inhibits the proliferation of oligodendrocyte progenitors [30] . To better understand the role of neuropsin in the regulation of demyelination , we quantified Cnp ( 2' , 3'-Cyclic-nucleotide 3'-phosphodiesterase ) and Mog ( Myelin Oligodendrocyte Glycoprotein ) expression from hippocampus and cortex derived mRNA after chronic oil or corticosterone injection ( n = 6 ) . There was no significant difference between WT and KO in both the hippocampus and cortex in the oil injected groups . However , in the corticosterone injected groups , whilst there was no difference in the cortex , there was more expression of Cnp ( p = 0 . 04 , t = 2 . 23 ) and Mog ( p = 0 . 03 , t = 2 . 4 ) in the hippocampus of KO mice ( Fig 4C–4F ) . Collectively , these data demonstrate that mice lacking neuropsin feature a protective effect on neurogenesis , dendritic morphology , and demyelination in the hippocampus after chronic elevated plasma corticosterone . The expression of c-Fos is commonly used as a neuronal activity marker following stimulus [31] . In order to better understand the neuronal activity in WT and KO hippocampi in response to high levels of corticosterone , we measured neuronal activation by detecting the number of c-Fos positive cells twenty-four hours after the last injection of corticosterone or oil ( 28 days of injections ) in both WT and KO mice ( n = 6 in each group ) . In the oil treated groups , our results show there was no difference between WT and KO mice . Interestingly , after corticosterone injection , WT mice had more c-Fos positive cells in the dentate gyrus ( DG ) ( p = 0 . 02 , t = 2 . 7 ) and CA3 ( p = 0 . 001 , t = 4 . 3 ) , but not in the CA1 ( p = 0 . 31 . t = 1 . 06 ) . Meanwhile , KO/CORT mice featured significantly fewer c-Fos positive cells in CA1 ( p = 0 . 03 , t = 2 . 43 ) and CA3 ( p < 0 . 0001 , t = 8 . 87 ) , but no significant difference in DG ( p = 0 . 33 , t = 1 . 01 ) compared to WT/CORT mice ( Fig 5 ) . These results demonstrate that KO mice exhibit a reduced response to corticosterone treatment as measured by hippocampal c-Fos expression . One of the major limitations of using constitutive KO animals is that aberrant development may potentially influence data . The KO model may have an effect on somatic cells and brain regions other than the hippocampus . To address this , we overexpressed neuropsin in the hippocampi of KO mice via an AAV9-CB-neuropsin viral vector ( NP/OE , n = 8 ) together with an AAV9-CB-GFP control ( NP/KO , n = 8 ) . Two weeks after surgery , we started to inject mice with corticosterone ( 20 mg/kg/day ) . After 14 days of corticosterone injections , we subjected the mice to a forced swim test , novelty suppressed feeding test and a sucrose preference test . Interestingly , NP/OE mice have longer immobility times in the forced swim test ( p = 0 . 003 , t = 3 . 62 ) , and a greater latency to approach food pellets in the novelty suppressed feeding test ( p = 0 . 018 , t = 2 . 72 ) in comparison with NP/KO mice . For the sucrose preference test , there was no significant difference between the two groups ( p = 0 . 09 , t = 1 . 84 ) ( S3A Fig ) . Our data shows that overexpression of neuropsin with concordant corticosterone injections leads to depressive-like behaviour . Next , we performed a water maze test to measure spatial learning and memory . There was no significant difference between the learning curves of both groups ( p = 0 . 99 , F11 , 66 = 0 . 0004 ) . We performed a memory test two weeks after training . The results showed no significant difference spent in the target zone between groups ( p = 0 . 172 , t = 1 . 45 ) but the number of times NP/OE mice crossed the phantom platform was significantly fewer ( p = 0 . 03 , t = 2 . 33 ) . This suggests overexpression of neuropsin may further impair spatial memory after chronic corticosterone injection ( S3B Fig ) . We confirmed hippocampal neuropsin overexpression by using qPCR . NP/OE mice featured a dramatic increase in neuropsin mRNA compared with NP/KO mice and NP/WT mice ( one- way ANOVA , p = 0 . 0006 , F2 , 12 = 14 . 96 ) ( S4B Fig ) . In contrast to the overexpression group , we performed surgery on wild type C57BL/6 mice ( n = 9 ) and injected either scramble sequence control or lentiviral vector shRNA to knockdown neuropsin in the hippocampus . Two weeks after surgery , we injected corticosterone daily on all mice . After 14 days of corticosterone injection , we evaluated the mice using the behaviour tests described above . Mice injected with the control scrambled viral vector featured a longer immobility time in the forced swim test ( p < 0 . 0001 , t = 6 . 34 ) and took a longer time to approach a food pellet in the novelty suppressed feeding test ( p = 0 . 01 , t = 2 . 74 ) . There was no significant difference between control and knockdown in the sucrose preference test ( p = 0 . 06 , t = 2 . 04 ) ( S3C Fig ) . Overall , these data suggest that inactivation of neuropsin protects against depressive-like behaviour . Next , we performed a water maze test to evaluate spacial learning and memory . The results showed no significant difference in the learning curves between groups ( p = 0 . 06 , F11 , 88 = 4 . 49 ) . Again , we performed the memory test two weeks after training . The results were consistent with the overexpression group: knockdown mice made significantly more platform crosses than the control group ( p = 0 . 02 , t = 2 . 39 ) but there was no significant difference in time spent within the target zone ( p = 0 . 18 , t = 1 . 38 ) ( S3D Fig ) . The expression analysis showed shRNA lentiviral vector significantly decrease neuropsin expression compared to the scramble control ( p < 0 . 0001 , t = 7 . 73 ) ( S4C Fig ) . To further investigate the mechanism of how neuropsin interacts with corticosterone to affect animal behaviour and hippocampal architecture , we analysed hippocampal gene and protein expression in four groups ( n = 6 in each group ) : WT/oil , KO/oil , WT/CORT , and KO/CORT . We first measured the hippocampal gene expression of neuropsin ( Klk8 ) . As expected , Klk8 expression was not detectable in KO mice . However , Klk8 expression was increased after chronic corticosterone injection in WT mice ( p = 0 . 03 , t = 2 . 46 ) . For GR ( glucocorticoid receptor ) , KO/CORT mice had higher expression compared to WT/CORT mice ( p = 0 . 04 , t = 2 . 3 ) . This change can also be detected at the protein level ( p = 0 . 03 , t = 2 . 45 ) ( Fig 6A and 6D ) . Next , we analysed GR downstream target genes: serum and glucocorticoid-regulated kinase 1 ( Sgk1 ) and FK506 binding protein 5 ( Fkbp5 ) . Sgk1 is thought to transfer N-methyl-D-aspartate ( NMDA ) receptors to the plasma membrane and have a regulating function in learning and memory and also mediate the effect of glucocorticoids on neurogenesis and brain function including learning and memory [32–34] . Fkbp5 is highly associated with depression and other psychiatric disorders and is thought to reduce ligand sensitivity of the GR and be involved in stress response [35 , 36] . KO/CORT mice showed a lower level of Sgk1 expression compared to WT/CORT mice ( p = 0 . 0004 , t = 5 . 25 ) and no difference in Fkbp5 ( p = 0 . 25 , t = 1 . 27 ) ( Fig 6A , 6E and 6F ) . With regard to protein levels , western blot results show no differences in both SGK1 and Fkbp5 protein expression . These results provide evidence that mice lacking neuropsin have a reduced response to elevated plasma corticosterone . It has been suggested that the imbalance of glutamate transmission is pathogenic in mood disorders [37] . Therefore , we measured the expression of regulatory genes that are involved in the removal ( reuptake ) of glutamate from the neuronal synaptic cleft in the hippocampus . We show that corticosterone injection reduces the expression of excitatory amino-acid transporter 1 ( EAAT1 ) gene expression ( F1 , 20 = 15 . 63 , p = 0 . 0008 ) . In corticosterone injected mice , KO mice have a higher level of EAAT1 gene expression ( p = 0 . 008 , t = 3 . 23 ) than WT controls . This difference between genotypes is also observed at the protein level ( p = 0 . 027 , t = 2 . 59 ) ( Fig 6B and 6I ) . Expression of the vesicular glutamate transporter , VGluT1 , is increased by corticosterone ( F1 , 20 = 60 . 72 , p < 0 . 0001 ) . There was no difference in VGluT1 mRNA expression between KO/CORT and WT/CORT ( Fig 6B ) . These data suggest that chronic corticosterone treatment reduces the efficiency of glutamate reuptake and that neuropsin inactivation partially rescues this defect . Emerging evidence suggests NMDA receptor dysregulation is implicated in stress induced depressive-like behaviours [38] . To better understand the interaction of neuropsin and corticosterone in regulating NMDA receptors , we measured different subtypes of NMDA receptors in the hippocampus , including 2A ( GluN2A ) and 2B ( GluN2B ) . KO/CORT mice had significantly lower expression of GluN2B compared to WT/CORT ( p = 0 . 006 , t = 3 . 44 ) but no significant difference in GluN2A . The protein expression of NR2B showed a similar pattern to gene expression but there was no significant difference between WT/CORT and KO/CORT . We also detected a lower level of Camk2b gene ( p = 0 . 014 , t = 2 . 93 ) and CamKII protein ( p = 0 . 036 , t = 2 . 41 ) expression in KO/CORT mice compared to WT/CORT mice but no difference in Camk2a expression ( Fig 6B , 6G and 6H ) . These results suggest that neuropsin inactivation may prevent increased calcium influx seemingly induced by corticosterone injection . Previous studies show that activation of NMDA receptors induces reactive oxygen species ( ROS ) production in neurons [39] . Evidence also suggests high levels of plasma corticosterone triggers ROS production leading to oxidative damage in the hippocampus and impairment of memory retention [40] . To test this hypothesis , we conducted a NADPH oxidase-dependent ROS test . In the oil treated groups , our results show no difference in superoxide anion production between WT/oil ( n = 13 ) and KO/oil ( n = 15 ) mice . However , KO/CORT mice ( n = 9 ) have lower superoxide anion in the hippocampus compared to WT/CORT ( n = 10 ) induced by NADPH ( p = 0 . 013 , t = 2 . 75 ) ( Fig 6C ) . This suggests that neuropsin inactivation reduces ROS generation resulting from prolonged exposure to a high level of plasma corticosterone .
In this study , we set out to investigate the role of neuropsin in depressive-like behaviour following chronic stress . We observed significant differences in depressive-like behaviour between neuropsin KO and WT mice . However these differences were only evident after animals had experienced a chronic stress regime . Neuropsin is highly expressed in the hippocampus; a region with high neuronal plasticity that has long been considered to be involved in stress , depression , learning and memory . From previous studies , it is unclear whether neuropsin plays a role in learning and memory . One report suggested that neuropsin knockout mice exhibit impairment in the water maze test and that neuropsin is necessary for establishment of LTP , thereby playing a significant role in memory acquisition [20] . However , another study shows no difference in water maze behaviour between KO and WT mice [21] . In our study , intriguingly , we did not detect any difference in learning and memory between WT and KO mice prior to chronic stress . However , after experiencing chronic stress , KO mice performed better in memory tasks than WT mice . Chronic stress leads to disorders such as depression through prolonged elevated glucocorticoids and it is well known that depression may alter the function of learning and memory [17] . From our previous study , we showed that a combination of different stressors leads to depressive-like behaviours in mice [41] . In this study , we used a combination of restraint , foot shock , and sleep deprivation stressors . We confirmed that this array of stressors elicited an increase in plasma corticosterone . We also confirmed that there was no significant difference in plasma corticosterone level between WT and KO groups at each stage of the experiment: before , during and after stress . Our data show that inactivation of neuropsin protects mice from developing depressive-like behaviours and limited the impairment of memory following stress , possibly due to the interaction between neuropsin and corticosterone . Neuropsin has been shown to be involved in anxiety following acute stress by regulating the dynamics of the EphB2-NMDA-receptor interaction [6] . This suggests there may be an interaction between neuropsin and corticosterone in emotional response . In order to investigate the interaction between neuropsin and corticosterone on depressive-like behaviour and memory , we mimicked one aspect of the hormonal state observed during prolonged stress by injecting mice with a high dose of corticosterone . After chronic corticosterone injection , we observed similar effects to that observed in the chronically stressed animals . Our results suggest that the effects of stress in neuropsin-deficient mice on emotional behaviours and memory performance may be mediated by elevated plasma corticosterone . Additionally , it is also well known that stress or elevated corticosterone can result in excitotoxicity and the modification of hippocampal architecture including decreased dendritic spine density , neurogenesis , and myelination [14 , 15 , 42 , 43] . Our data suggest neuropsin deficient mice are protected against the effects of prolonged elevated corticosterone . This is supported by an assessment of neuronal activity as measured by c-Fos expression . Previous evidence demonstrated elevated plasma corticosterone can be detected more than twelve hours after subcutaneous corticosterone injection [44] . Consequently , we measured c-Fos expression twenty-four hours after the last corticosterone injection in order to assess neuronal activity following our chronic corticosterone treatment . KO mice featured reduced expression of c-Fos in CA1 and CA3 , an area where neuropsin is highly expressed in WT after chronic corticosterone treatment . This suggests that inactivation of neuropsin prevents the occurrence of neuronal hyperactivity following chronic corticosterone injection . This may explain why neuropsin KO mice exhibit milder behavioural defects after chronic corticosterone injection than WT . By intra-hippocampal injection of viral vectors , we are able to understand the function of neuropsin more specifically within the hippocampus . In our study , we used a lentiviral vector to knockdown neuropsin on C57BL/6 mice . Additionally we used an AAV9 vector to overexpress neuropsin in the hippocampus of KO mice . Our data support that the development of depressive-like behaviour following chronic corticosterone injection is mediated by the presence of neuropsin . Results from gene expression and protein quantification of GR reveal KO mice exhibit a reduced response to elevated corticosterone . This can also be confirmed by Sgk1 gene expression . Sgk1 is a direct GR target gene and has been shown to inhibit neurogenesis and be increased in expression in unmedicated depressed patients [33] . A previous study has also shown that Sgk1 up-regulates the expression of NMDA receptor subunit GluN2A and GluN2B [34] . Interestingly , we detected higher levels of gene expression for GluN2B in WT/CORT mice compared to KO/CORT mice , although no significant difference in protein level was observed via western blot . Previous studies have shown GluN2B receptors regulate depressive-like behaviours and mediate excitotoxic neuronal damage [45 , 46] . Moreover , WT/CORT mice also exhibited higher levels of CamkII , indicating increased levels of intracellular calcium [47] . WT/CORT mice also have lower expression of EAAT1 . This may lead to synaptic glutamate accumulation . Excess glutamate at synapses and an increased influx of calcium ions may trigger downstream responses that increase ROS production [39] . This hypothesis is supported by our ROS experiment . WT/CORT mice display higher level of NADPH dependent ROS in the hippocampus compared to KO/CORT mice . It has been suggested that imbalance between the production of free radicals and the antioxidant capacity of an organism , may contribute to the neuropathology of neurological and psychiatric diseases , including major depression [48] . In our study , repeated corticosterone injection increases ROS production capacity in the hippocampus . This may contribute to the neuronal and behavioural alterations that we have observed . Our data suggests neuropsin inactivation protects against corticosterone induced impairments in hippocampal neuronal architecture and the development of depressive-like behaviours . The lower ROS production capacity of the KO mice chronically injected with corticosterone may help explain the protective effects we observe with this model . In conclusion , we have identified a novel link between the serine protease neuropsin , chronic stress and depression . We show that neuropsin inactivation has protective effects against depressive-like behaviours and partially rescues spatial memory after stress . These effects could be due to an interaction between elevated corticosterone and neuropsin . Overall , we provide a gene–environment interaction study suggesting that blockade of neuropsin activity reduces the impairment caused by chronic stress or elevated plasma corticosterone . This pathway may consequently be a strong candidate for clinical interventions in the treatment of stress disorders .
Experiments were performed on eight-week-old littermate wild-type and neuropsin knockout mice [49] with an equal mixed gender balance . Mice were maintained in specific pathogen-free conditions . They were housed in a 12:12 hour light dark cycle at a temperature of 22°C and a humidity level of 60–70% . Animals had ad libitum access to food and water . All the protocols to proceed in this study were reviewed and approved by the Institutional Animal Care and Use Committee at Chang Gung University ( Permit Number: CGU14-060 ) . The stress procedure combined restraint stress , foot shock , and sleep deprivation . Animals were subjected to 3 hours restraint stress on the first day . The following day they were placed in a plexi glass foot shock chamber and given two inescapable 1 mA foot shocks lasting 5 seconds within a period of 3 minutes . The next day , mice were subjected to sleep deprivation by being placed in a water tank containing multiple and visible platforms ( 4 . 5 cm in height and diameter ) surrounded by water for 12 hours [50] . This battery of stress inducing paradigms was repeated for 28 days . Novelty suppressed feeding: Food pellets were removed from the home cage 24 hours before testing . Mice were placed for 7 min in a novel cage with food pellets placed in the 4 corners and a bright light positioned above to illuminate the whole testing chamber . All activity was tracked by a video tracker . Both the latency to initiate eating and the time spent eating were measured . Forced swim test: Mice were put into a water tank ( diameter = 20 cm ) , containing 15 cm of water for 6 min and their immobility time was measured . Morris water maze: The water maze has four starting positions: north , south , east , and west . Before beginning the test we chose the order of the starting directions for consistency: north , east , south and then west . For training , mice were placed in the maze and allowed to swim/search for the hidden platform for a maximum of 90 seconds . The time taken to reach the platform was recorded . If the mice did not find the platform within 90 seconds they were guided to the platform . This procedure was repeated three times for each starting direction over three days ( 12 trials in total ) . We then performed the memory test 14 days after training was complete , with the platform removed . We recorded the time each mouse spent in the correct quadrant ( target zone ) where the training platform was previously located . We also recorded how many times the mice crossed the phantom platform location . Sucrose preference: For a habituation period of three days , mice were presented with two drinking bottles; 1% sucrose solution and drinking water . The position of the two bottles was switched every day to reduce side bias . Following habituation , mice were not given water for 6 hours prior to testing . We measured the net weight of both water ( W0 ) and sucrose ( S0 ) before the test and 24 hours later ( W1 , S1 ) when the test was completed . The preference ( sucrose consumption % ) was calculated by the following formula:Sucrose consumption ( % ) = ( S0−S1 ) / [ ( S0−S1 ) + ( W0−W1 ) ]*100 . Blood samples were collected to measure basal corticosterone level and the level after 3 hrs of restraint stress , 30 min after footshock , after 6 hrs of sleep deprivation and 24 hrs after all stressors via facial vein puncture . Plasma was separated from whole blood by centrifugation ( 3000g , 4°C , 10 min ) and stored at -80°C until used . Plasma was diluted 1:30 in buffer and measured using the Corticosterone EIA kit ( Enzo Life Sciences ) according to the manufacturer’s instructions . Brains were harvested and fixed in 4% PFA ( pH7 . 4 ) for a day . The brains were then dehydrated for 24 hrs in 25% sucrose solution . All sections for KI67 , DCX , c-Fos and CNPase staining were cut to a thickness of 30 μm on a sliding microtome . Sections were mounted on superfrost slides and dried overnight . Subsequently , slides were incubated in 0 . 01 mol/L citric buffer for 20 min at 90°C , 3% H2O2 for 10 min , rinsed in PBS , and incubated overnight at room temperature in rabbit anti-KI67 antibody ( 1:4000 , Vector Lab ) , goat anti-DCX antibody ( 1:250 , Santa Cruz ) , rabbit anti-c-Fos ( 1:1000 , Santa Cruz ) , or rabbit anti-CNPase ( 1:1000 , Abcam ) . The next day , a standard IgG ABC kit ( Vector Lab ) procedure was used and the slides reacted for 5–10 min with a Sigma DAB tablet . Sections were then counterstained with cresyl violet and cover-slipped with DPX . For BrdU staining , following a 3% H2O2 incubation for 10 min , slides were subsequently incubated in 2M HCL for 30 min at 37°C . Rat anti-BrdU antibody ( 1:250 , Accurate ) was applied overnight . The next day the ABC kit procedure was followed and the slides were reacted with a Sigma DAB tablet . Hippocampus cells were counted bilaterally on every eighth section through the entire rostrocaudal extent of the granule cell layer . All slides were randomized and coded before quantitative analysis . Slides ( half brain ) were examined under a 20× objective . DCX , KI67 and BrdU labelled cells were counted on every eighth section through the entire rostrocaudal extent of the granule cell layer ( 6 sections per animal ) . The number of cells counted was then multiplied by sixteen to obtain an estimate of the total number of DCX , KI67 and BrdU positive cell in the dentate gyrus . For c-Fos labelled cells , we counted cells on every eighth section through the entire rostrocaudal extent of the dentate gyrus , CA1 and CA3 areas ( 6 sections per animal ) . The raw data was then modified as detailed above . Brains were harvested and stained using the superGolgi Kit ( Bioenno Tech ) in accordance with the manufacturer's instructions . Brains were impregnated with solution for 9 days . This was followed by 2 days of dehydration in sucrose . Brains were subsequently sectioned coronally at a thickness of 150 μm using a microtome . We calculated the spine density of the secondary branches of pyramidal neuron apical dendrites in CA3 . We selected a total of 18 neurons per each animal . Three different neurons were randomly selected for measuring per brain slice in both WT and KO mice ( n = 4 ) . A total of six brain slices were analysed per animal . For the neuropsin overexpression experiment and control , DNA fragments that encoded KLK8 or EGFP were created by PCR and subcloned into the NotI site of a AAV9 virus construct . Recombinant AAV9 vectors were produced by a standard triple-plasmid transfection method and purified by two rounds of CsCl centrifugation . The physical vector titres of AAVs were quantified by measuring the number of packaged vector genomes using a real-time PCR method . For the neuropsin knockdown experiment and control , a klk8 short hairpin RNA ( shRNA ) lentiviral vector and scrambled sequence control vector were obtained from the National RNAi core facility ( Institute of Molecular Biology , Academia Sinica , Taiwan ) . The target sequence for Klk8 shRNA was: CCTCAACTGTGCGGAAGTGAA . For the viral vector injections , surgery was performed under anaesthesia ( a Zoletil-Rompun mixture ) . Mice received bilateral injections of viral vectors into hippocampus in a volume of 1 . 5 μl ( 5x1013 VG/ml for AAV9 viral vector , 5 . 6x106 RIU/ml for lentiviral vector ) under stereotaxic guidance . There were four injection sites in total ( For dorsal hippocampus , AP– 2 . 2mm , ML ± 2 . 0 mm from bregma and DV—1 . 8 mm from dura; for ventral hippocampus , AP—3 . 0mm , ML ± 3 . 0 and DV—3 . 3 mm ) . ( S4A Fig ) Hippocampal tissue was collected from WT and KO mice . mRNA was extracted using a RNeasy Lipid Tissue Kit ( Qiagen ) . cDNA was then made using SuperScript III reverse transcriptase ( Invitrogen ) . Experiments were performed in duplicate . Gene expression levels were then calculated by the ΔΔCt method and normalized against a Gapdh control . Primers used in this study are in the Table 1 . Protein extracts were obtained by lysing tissue in 1% SDS . Samples were homogenized and heated at 100°C for 10 min . Proteins were then separated by sodium dodecyl sulphate–polyacrylamide gel electrophoresis and electro-transferred onto nitrocellulose membranes . Blots were incubated with primary antibody ( GR 1:1000 , Santa Cruz; SGK1 1:1000 , BOSTER; Fkbp5 1:3000 , Santa Cruz; NR2b 1:1000 , Millipore; CamKII 1:1000 , Santa Cruz; EAATI 1:1000 , Alpha Diagnostic ) placed in TBST and 5% non-fat milk overnight at 4°C . Subsequently , blots were washed and probed with the respective horseradish peroxidase secondary antibody for 1 h at room temperature . The immunoreactive bands were visualized using ECL detection reagent ( GE Healthcare , RPN2106 ) . Assessment of the band intensities were performed using the ChemiDoc MP from Bio-Rad . Superoxide anion production in the hippocampus tissues was measured by modified lucigenin-enhanced chemiluminescence . This method to test superoxide anion production level in tissue is based on the nitro blue tetrazolium ( NBT ) reduction method [51] . Hippocampi were isolated , chopped and then placed in a white plate containing Krebs buffer . Lucigenin ( 25 μM ) was added to test the basal level of superoxide anion production . NADPH ( 500 μM ) was then added to stimulate NADPH-oxidase dependent ROS production . The magnitude of superoxide anion production was normalized for the dry weight of the chopped hippocampus . The mean ± SEM was determined for each group . Statistical analyses were performed using Graphpad Prism software . Data were analyzed via an analysis of variance ( ANOVA ) or t-test as appropriate . Fisher's LSD method was performed when applicable . Differences were considered significant when p was less than 0 . 05 . | Depression is a medical condition that results in significant morbidity , mortality and reduced quality of life . Understanding the molecular mechanism in which stress leads to depression is essential for the discovery of new clinical interventions . Currently , despite considerable research , the mechanisms underlying stress-related illnesses are unclear . In this study , we reveal a novel link between the interaction of serine protease neuropsin with corticosterone and the development of chronic stress induced depressive-like behaviour . We found no difference in baseline behaviour between neuropsin knockout and wild-type mice . However , our results show that neuropsin knockout mice are protected against the development of depressive-like behaviours and memory impairment following chronic stress . We suggest the dysregulation of the glutamate system may lead to increased reactive oxygen species and act as a possible mechanism in which stress changes hippocampal architecture . We found that inactivation of neuropsin limits the extent of these anatomical changes in both chronic stress and corticosterone injection exposed cohorts . In this study , we outline a mechanism that may open new possibilities for the treatment of stress-related psychiatric disorders . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"medicine",
"and",
"health",
"sciences",
"viral",
"transmission",
"and",
"infection",
"brain",
"social",
"sciences",
"vertebrates",
"mice",
"neuroscience",
"animals",
"mammals",
"learning",
"and",
"memory",
"microbiology",
"viral",
"vectors",
"animal",
"behavior",
"cognition",
"memory",
"zoology",
"mood",
"disorders",
"behavior",
"epidemiology",
"mental",
"health",
"and",
"psychiatry",
"disease",
"vectors",
"rodents",
"hippocampus",
"psychology",
"anatomy",
"virology",
"biology",
"and",
"life",
"sciences",
"depression",
"cognitive",
"science",
"amniotes",
"organisms",
"psychological",
"stress"
] | 2016 | Neuropsin Inactivation Has Protective Effects against Depressive-Like Behaviours and Memory Impairment Induced by Chronic Stress |
Amino acid covariation , where the identities of amino acids at different sequence positions are correlated , is a hallmark of naturally occurring proteins . This covariation can arise from multiple factors , including selective pressures for maintaining protein structure , requirements imposed by a specific function , or from phylogenetic sampling bias . Here we employed flexible backbone computational protein design to quantify the extent to which protein structure has constrained amino acid covariation for 40 diverse protein domains . We find significant similarities between the amino acid covariation in alignments of natural protein sequences and sequences optimized for their structures by computational protein design methods . These results indicate that the structural constraints imposed by protein architecture play a dominant role in shaping amino acid covariation and that computational protein design methods can capture these effects . We also find that the similarity between natural and designed covariation is sensitive to the magnitude and mechanism of backbone flexibility used in computational protein design . Our results thus highlight the necessity of including backbone flexibility to correctly model precise details of correlated amino acid changes and give insights into the pressures underlying these correlations .
Evolutionary selective pressures on protein structure and function have shaped the sequences of today's naturally occurring proteins [1]–[3] . As a result of these pressures , sequences of natural proteins are close to optimal for their structures [4] . Natural protein sequences therefore provide an excellent test for computational protein design methods , where the goal is to predict protein sequences that are optimal for a desired protein structure and function [5] . It is often assumed that given a natural polypeptide backbone conformation , an accurate protein design algorithm should be able to predict sequences that are similar to the natural protein sequence . This test is commonly referred to as native sequence recovery [4] and it has been used extensively to evaluate various protein design sampling methods and energy functions [6]–[8] . Beyond simply recovering the native sequence , a further challenge in computational protein design is to predict the set of tolerated sequences that are compatible with a given protein fold and function [9]–[13] . Predicting sequence tolerance is important for applications such as characterizing mutational robustness [14] , [15] , predicting the specificity of molecular interactions [16]–[20] , and designing libraries of proteins with altered functions [21] , [22] . Recent methods developed for this goal involve generating an ensemble of backbone structures similar to the native structure and then designing low energy sequences for the different structures in the ensemble [9] , [16] , [19] , [23]–[25] . These flexible backbone design methods can produce sequences that are highly divergent from the native sequence but may still fold into the desired structure , which makes simple native sequence recovery a poor indicator for the accuracy of these methods . A more useful computational test of these approaches involves comparing designed sequences with a set of reference sequences , either naturally occurring or experimentally derived , that share the desired protein fold . This comparison can be based on sequence profile similarity , which involves quantifying the difference between the frequencies of observing each amino acid at corresponding positions in the designed and reference sequences [16] , [17] , [19] . While high similarity between designed and reference sequence profiles can be informative to gauge the accuracy of a protein design method , it does not guarantee that the method will predict sequences that fold into the desired structure . This is because sequence profile comparisons evaluate amino acid positions independently from each other and therefore ignore the details of amino acid interactions that are critical for protein structure and function . Naturally occurring protein structures are formed cooperatively and each amino acid can physically interact with multiple neighboring amino acids . Evolutionary selective pressures have acted upon these interactions , resulting in the patterns of amino acid covariation that can be observed within today's naturally occurring protein families . Accordingly , previous studies have used information theoretic methods to detect amino acid covariation in multiple sequence alignments of many different protein families [26]–[28] and have used contact prediction based on covariation to dramatically improve the accuracy of protein structure modeling [29] . Despite the clear occurrence of amino acid covariation in natural protein sequences , the extent to which different selective pressures have shaped amino acid covariation in diverse protein families is unknown . Additionally , it is difficult to dissect to what extent phylogenetic bias has influenced the observations of amino acid covariation . Previous work has indicated that networks of covarying amino acids play a role in allosterically linking distant functional sites , suggesting that amino acid covariation is driven by protein functional constraints [30] , [31] . However , other studies have shown in two test cases that computational protein design can recapitulate naturally occurring covariation in the cores of SH3 domains [4] , [13] , [32] and for two-component signaling systems [33] . These results indicate that constraints imposed by protein structure have played a role in producing the covariation in the studied examples , but it has not yet been shown that these observations are general . In this paper , we use computational protein design to measure the extent to which protein structure has shaped amino acid covariation in a diverse set of 40 protein domains . Since computational protein design predicts sequences that are energetically optimal based on protein structure alone , we expect that pairs of amino acids that highly covary in both designed and natural sequences to have likely covaried to maintain protein structure . We find significant overlap in the sets of highly covarying amino acid pairs between designed and natural sequences for all 40 domains examined , suggesting that maintenance of protein structure is a dominant selective pressure that constrains the evolution of amino acid interactions in proteins . Our analysis furthermore quantifies the extent to which different types of interactions explain the observed covariation . Finally , we demonstrate the utility of amino acid covariation recapitulation as a sensitive test for evaluating different protein design methods . We find that flexible backbone design significantly improves covariation recapitulation relative to fixed backbone design and that recapitulation of amino acid covariation is exquisitely sensitive to both the magnitude and mechanism of backbone flexibility . Taken together , these results provide fundamental insights into the physical nature of amino acid co-evolution and , more practically , provide a new benchmark that may help improve the accuracy of computational protein design methods .
To compare amino acid covariation in natural and predicted designed protein sequences , we selected 40 protein domains that were diverse with respect to their secondary structure composition and fold class ( Table 1 ) . We then quantified natural amino acid covariation for each domain by creating a multiple sequence alignment for the domain , followed by computing covariation between every pair of columns in the multiple sequence alignment by using a mutual information based method [28] ( see Methods ) . Pairs of amino acid positions with a covariation score that is two standard deviations above the mean or greater were considered to be highly covarying pairs . We predicted designed protein sequences for each of the 40 domains using RosettaDesign [4] , [34] . We first used the standard RosettaDesign fixed backbone protocol [34] , which takes a crystal structure as input and runs Monte Carlo simulated annealing , to predict 500 designed sequences for each domain structure . We then quantified amino acid covariation in the designed sequences and compared it to natural amino acid covariation for each domain . We calculated the similarity between designed and natural covariation based on the percent overlap of the highly covarying pairs in each set ( see Methods ) . We found this overlap to be significant ( p<0 . 001 ) for all 40 domains ( Table S1 ) . Given the observation that fixed backbone protein design can recapitulate a significant fraction of naturally covarying amino acid pairs , we next aimed to understand how incorporating backbone flexibility into the design protocol affects this recapitulation . To accomplish this , we generated a conformational ensemble of 500 backbone structures for each domain using the “backrub” method [35] in Rosetta [36] , which iteratively applies local backbone perturbations throughout the protein structure combined with adjustments in side-chain conformations . We then used RosettaDesign to predict a low energy sequence for each backbone structure in the ensemble , resulting in 500 designed sequences . Figure 1 shows a flow chart of this approach applied to an SH3 domain . To investigate the effect of the magnitude of backbone flexibility in the design protocol , we varied the temperature parameter in the Monte Carlo simulations used in the backrub protocol to generate conformational ensembles with different amounts of structural variation ( Figure 2A ) . We designed sequences for each ensemble ( kT = 0 . 3 , 0 . 6 , 0 . 9 , 1 . 2 , 1 . 8 , 2 . 4 ) and quantified similarity to natural covariation for each set of sequences . We compared these results with sequences designed using the fixed backbone design protocol described above ( “Fixed” ) . Figure 2B shows a significant increase in covariation similarity for the flexible backbone simulations relative to the fixed backbone simulation . Moreover , the distributions of covariation similarity for the 40 domains show that there is an optimal degree of structural variation , as low-temperature and high-temperature simulations perform significantly worse than mid-temperature simulations ( Table S2 ) . We observed this same trend when we repeated this analysis using a different method for quantifying covariation [37] ( Figure S1 ) , suggesting that our results are not dependent on the method used to quantify covariation . To better understand the basis of this trend , we examined several other sequence and structural characteristics: sequence recovery , sequence profile similarity , sequence entropy and structural variation ( see Methods ) . The resulting distributions for these characteristics are shown in Figure 2C . Sequence entropy and sequence profile similarity showed similar trends to covariation similarity ( sequence entropy is most similar to natural sequences and profile similarity is highest at 0 . 9 kT ) , suggesting that backbone flexibility allows for sampling diverse sequences with native-like properties . These trends are consistent with the observation that sequence recovery decreases with increasing amounts of backbone flexibility . As diversity within a set of sequences increases , those sequences tend to become more dissimilar to any individual sequence , including the native sequence of the crystal structure used as input for design . Structural variation in the 0 . 3 , 0 . 6 , 0 . 9 and 1 . 2 kT simulations is less than the structural variation among naturally occurring protein structures with these domains , which could be due to the fact that natural proteins use additional mechanisms of generating structural variation that are not being modeled , such as the insertion or deletion of amino acids in loop regions . Taken together , these results suggest that a moderate degree of backbone flexibility allows for the accommodation of sequences that differ from the native sequence and yet are similar to naturally occurring sequences with respect to their sequence profiles , sequence entropies and patterns of amino acid covariation . Next we examined whether or not these results were specific to the method used to generate the conformational ensembles for design . We tested two other Monte Carlo based methods that iteratively perform perturbations to the backbone . One method performed Kinematic Closure ( “KIC” ) , which involves randomizing phi/psi torsions in a local region of the backbone while keeping the rest of the backbone fixed , thus introducing a chain break , and then using inverse kinematics to solve for the torsions that will close the chain [38] . The other method performs potentially non-local moves by perturbing the phi and psi torsions of residues by a random small angle ( “Small” ) [39] . We ran both of these methods for the same number of trials and for the same values of kT as the backrub protocol . The resulting distributions of covariation similarity show the same trend we observed previously with the backrub simulations , where mid-range temperature simulations result in an optimal degree of covariation similarity ( Figure S2 ) . While the optimal simulation temperature parameter was comparable for each of the methods tested , the methods achieved a different optimum level of covariation similarity with the natural sequences . We found that the two local move simulations ( KIC and Backrub ) outperformed the non-local move simulation ( Small ) . To test if this observation holds true more generally , we tested two additional methods of generating conformational ensembles that make non-local moves . These methods included FastRelax ( “Relax” ) , which consists of multiple rounds of side-chain repacking and all-atom minimization while increasing the weight of the repulsive term in the Lennard–Jones ( LJ ) potential from 2% to 100% of its default value , and AbInitioRelax ( “AbRelax” ) , which performs fragment-based ab initio structure prediction followed by FastRelax [40] . As an additional control , we also designed sequences using a fixed backbone structure with an energy function that dampens the weight of the repulsive LJ term ( “Soft” ) . The resulting covariation similarity distributions show that recapitulation of natural amino acid covariation is sensitive to the method used to generate conformational ensembles ( Figure 3A ) . Both local move simulations ( KIC , Backrub ) achieved higher median covariation similarities than the non-local move simulations ( Small , AbRelax , Relax ) and the fixed backbone simulations ( Fixed , Soft ) ( see Table S3 for p-values ) . We also evaluated each of these methods using the other metrics described above: native sequence recovery , sequence profile similarity , sequence entropy and structural variation ( Figure 3B ) . Unexpectedly , the AbRelax method , which resulted in conformational ensembles with the greatest structural variation , achieved the highest sequence profile similarity with the natural sequences of any method tested . A possible explanation for this behavior is that local interactions are preserved in AbRelax generated structures , but the overall topology of the protein is incorrect . To test this hypothesis , we examined covariation similarity in the AbRelax sequences by splitting all covarying pairs into the following two sets: pairs separated by fewer than 10 residues in sequence ( “Near” ) and pairs separated by greater than 10 residues in sequence ( “Far” ) . This analysis revealed that whereas AbRelax sequences have relatively high covariation similarity with natural sequences for pairs close in sequence , they have low covariation similarity for pairs that are distant in sequence ( Figure 3C ) . In contrast , covariation similarity for “near” and “far” pairs were similar for simulations using backrub ensembles . These results suggest that AbRelax can model local interactions within a secondary structural element or between adjacent secondary structures , but it does not correctly capture non-local interactions that are likely critical for achieving a cooperatively folded , stable tertiary structure . This observation demonstrates the importance of using amino acid covariation to evaluate the accuracy of protein design methods , since it is possible to obtain deceptively high sequence profile similarity scores with highly divergent tertiary structures as long as local interactions are maintained . Of all the flexible backbone design methods tested , Backrub , kT = 0 . 9 resulted in sequences most similar to the natural sequences with respect to covariation similarity and sequence profile similarity . Using the assumption that a method that gives higher similarity to natural sequences will better capture the mechanisms underlying covariation , we used Backrub , kT = 0 . 9 as the representative flexible backbone sequences for the remainder of the study . To understand how backbone flexibility influences the extent of covariation similarity between designed and natural sequences , we identified all pairs of amino acid positions that highly covaried in both the natural sequences and a set of flexible backbone sequences ( Backrub , kT = 0 . 9 ) but did not highly covary in the fixed backbone sequences . We then took all pairs of amino acids at these positions that were not sampled in the fixed backbone simulation and designed them onto the crystal structure backbone using fixed backbone design . For each pair of these positions , we calculated mean interaction energies and compared these energies between fixed and flexible backbone design structures ( Figure 4A ) . We calculated both one-body energies , which include the interaction of an amino acid residue with itself , and two-body energies , which include the interactions between two amino acid residues in the protein ( see Text S1 for description of the components of Rosetta one-body and two-body energies ) . We found both the one-body and two-body energies of these pairs to be generally greater in the context of fixed backbones relative to flexible backbones . Splitting the energies into their component terms revealed that the backbone-dependent Dunbrack rotamer energy ( fa_dun ) and Lennard-Jones repulsive ( fa_rep ) terms resulted in greater energy increases in the one-body and two-body energies , respectively , than any other term in the energy function ( Figure S3 ) . These results suggest that amino acid pairs that covary in flexible backbone simulations but do not covary in fixed backbone simulations generally cannot be accommodated on fixed backbones without resulting in steric clashes or rotamers that are unfavorable for the given backbone . Simply modifying the energy function by using a “soft” repulsive potential that reduces the energy of clashes does not increase sequence diversity or covariation similarity ( Figure 3B ) , suggesting that backbone movements are required to accommodate these amino acid interactions . Figure 4B shows representative cases where some degree of backbone flexibility is required to correctly model the precise interaction details of specific amino acid pairings . We have thus far compared amino acid covariation between natural and predicted designed sequences based on the extent of overlap between the sets of highly covarying pairs . However , it is also important to consider the amino acid pair propensities at covarying positions to test whether the natural and designed covarying pairs utilize the same types of amino acid interactions . To accomplish this , we calculated amino acid propensities at pairs of positions that covary in both the natural and designed sequences ( Figure 5A ) . Over-represented amino acid pairs in both designed and natural sequences included those with opposite charges , hydrophobic pairs and hydrogen-bonding pairs . Differences in the designed and natural amino acid pair propensities included the over-representation of cation-pi pairs in the natural sequences but not in the designed sequences ( such as W-R ) . These differences highlight shortcomings of the energy function used for design , which does not currently account for cation-pi interactions . To quantify the similarity between the natural and designed covarying pair propensities , we calculated the correlation coefficients between the natural and designed propensities for all sets of designed sequences . We found these correlations to be dependent on both the magnitude and mechanism of backbone flexibility , as we previously observed with the overlap in covarying pairs ( Table S4 ) . The comparison between natural and designed pair propensities for fixed backbone sequences and for a set of flexible backbone sequences ( Backrub , kT = 0 . 9 ) are shown in Figure 5B , again supporting the conclusion that backbone flexibility improves recapitulation of amino acid covariation . While similar pair propensities between natural and designed covarying pairs demonstrate that the same types of amino acid interactions occur in both natural and designed sequences , they do not show that the mechanisms underlying covariation are the same in both cases . To investigate this , we first classified the mechanism of covariation for all pairs that covary in both designed and natural sequences and then quantified how often the same mechanism is used . Figure 6A shows an illustration of three of the covariation mechanisms: size , hydrogen bonding and charge . Classifying each of these mechanisms requires examining the transition from one amino acid pair to another . For example , the transitions depicted in Figure 6A are IA–VV , AP–SS , RE–DR . Covariation due to size involves a decrease in the size of one amino acid and an increase in the size of the other ( IA–VV ) . Covariation due to hydrogen bonding involves a hydrogen bond that exists in one pair but not the other ( AP–SS ) . Covariation due to charge involves a pair of amino acids with opposite charges that either swap sign ( RE–DR ) or become uncharged amino acids . We also defined covariation mechanisms based on cation-pi interactions , pi-pi interactions , and other interactions not falling into any of the previous categories that we classify as hydrophobic , hydrophilic or mixed hydrophobic and hydrophilic ( see Methods for a detailed definition ) . For each pair of positions that covaried in both the designed and natural sequences , we computed the ten most significant transitions between amino acid pairs at those positions and classified each transition based on the mechanism of covariation . The resulting distributions of covariation mechanisms for the designed and natural pairs are shown in Figure 6B . The designed and natural covariation mechanisms distributions share similar properties , including covariation due to charge being the most common mechanism , whereas cation-pi , pi-pi and other ( hydrophilic ) covariation mechanisms are more rare . In both natural and designed distributions , hydrogen bonding and size covariation together account for approximately 30% of the total mechanisms . However , a number of quantitative differences exist in the distributions , including charge occurring more frequently in the designed pairs , suggesting that the design method may be over-predicting charged interactions . Additionally , in the natural pairs , size covariation is more common than hydrogen bonding covariation while the opposite is true in designed pairs . The “other” categories are also more common in the natural pairs than in the designed pairs . To better understand these differences , we split the pairs up based on the extent of their burial and compared the distributions of covariation mechanisms ( Figure S4 ) . This analysis revealed that covariation mechanism is dependent on the extent of pair burial and that buried pairs have the most significant differences between natural and designed covariation mechanisms . In natural buried pairs , the most common covariation mechanisms are size and other ( hydrophobic ) , whereas the most common mechanisms in designed buried pairs are hydrogen bonding and size . This likely occurs due to insufficient penalization of buried polar groups during the design protocol , resulting in over-predicting polar amino acids at buried positions and therefore incorrect predictions of covariation mechanism . To quantify how often the same covariation mechanism is used for specific pairs of positions in the designed and natural sequences , we calculated the percent of pairs sharing the same classification type in both the natural and designed sequences ( percent overlap ) for each type of covariation mechanism ( Figure 6C ) . Covariation due to charge has the highest percent overlap between the designed and natural pairs , followed by hydrogen bonding , size , other ( hydrophobic ) and other ( mixed ) , which have roughly equal percent overlaps . Covariation due to cation-pi and pi-pi interactions have relatively low percent overlaps between the designed and natural sequences , likely due to the fact that these types of interactions are not explicitly accounted for in the design energy function . We repeated this analysis using fixed backbone design sequences and found a decrease in the percent overlaps for size and other ( hydrophobic ) interactions , indicating that backbone flexibility may aid in modeling these types of covariation mechanisms ( Figure S5 ) . Taken together , this analysis provides insights into the mechanisms underlying amino acid covariation in naturally occurring proteins . Overall , the analysis shows considerable agreement between naturally occurring and designed covariation mechanisms . In some cases , it exposes pathologies in the design methods ( such as the over-representation of polar amino acids in cores under-representation of cation-pi and pi-pi interactions ) that can be addressed in future work using naturally occurring covariation as a reference point . While computational protein design can model a significant fraction of naturally occurring covarying amino acid pairs , there remain pairs of amino acids that are highly covarying in the natural sequences but not in the designed sequences ( nature-specific pairs ) . Moreover , there also exist pairs that highly covary in designed sequences but not in natural sequences ( design-specific pairs ) . Figure 7A shows the classification of nature-specific , design-specific and overlap pairs for the SH3 domain . To understand the basis for these differences , we first compared these sets of pairs based on their distances in three-dimensional structure ( Figure 7B ) . We found the design-specific and overlap covarying pairs to be significantly closer in structure than the nature-specific pairs . These results are consistent with the all-atom energy function used for generating the design sequences , which is most sensitive at short distances . The long distances in the nature-specific pairs could result from a number of factors , including interactions that bridge monomers in an oligomeric complex [37] , interactions that exist in alternative conformations [37] , long-range correlations in protein dynamics or from phylogenetic bias in the natural sequences . Another possibility is that in naturally occurring proteins , destabilizing substitutions ( that occur in functional sites ) co-vary with compensating stabilizing mutations in the protein that could be far away from the functional site . In addition to analyzing design-specific and nature-specific pairs with respect to pair distance , we compared them based on extent of amino acid burial , the presence in interfaces or active sites , and amino acid pair propensity . We observed a slight decrease in the percent of exposed pairs in the designed-specific pairs relative to the nature-specific pairs ( Figure S6 ) , which may be due to the difficulty of accurately modeling solvent exposed interactions in protein design . We observed no difference in the design-specific and nature-specific pairs with respect to their presence in interfaces or active sites ( Figure S7 ) , suggesting that the constraints imposed by known functional sites are not responsible for the inability to model the nature-specific pairs . We observed that the amino acid pair propensities of nature-specific and overlap pairs were different , while the amino acid pair propensities of design-specific pairs were highly correlated to those of the overlap pairs ( Figure 7C ) . The latter observation indicates that the energetic interactions leading to design-specific and overlap pairs may be similar to each other . A simple explanation may be that the design-specific pairs are equally compatible with the given protein structure , but may simply not have been sampled by nature . Such design-specific pairs may provide opportunities for engineering proteins with novel amino acid interactions , such as re-designing the specificity of protein-protein interactions .
Our study tested the hypothesis that the structural constraints imposed by protein architecture are a major determinant of amino acid covariation in naturally occurring proteins . If true , we reasoned that computational design methods that design sequences based on protein structure alone should be able to recapitulate amino acid covariation , provided that design predictions are sufficiently accurate . Confirming these ideas , we found a significant overlap between amino acid covariation in natural and designed protein sequences across a set of 40 diverse protein domains . These results quantify the influential role of the selective pressures for maintaining protein structure on shaping amino acid covariation . Therefore , even though correlated changes are undoubtedly important to evolve new activities and regulatory mechanisms [30] , [31] the presence of covariation alone may not necessarily indicate a functional role . Our study also illustrates how recapitulation of amino acid covariation serves as a stringent test for the ability of computational protein design methods to capture precise details of interactions between amino acids . We demonstrate that modeling backbone flexibility significantly increases the similarity between natural and designed covariation , and that this similarity is exquisitely sensitive to the mechanism used to model backbone changes . These findings indicate that protein backbone motions are required for allowing precise adjustments in amino acid interactions that enable covariation . Moreover , simulations that perform local backbone movements ( Backrub and KIC ) result in sequences with more natural-like covariation than simulations that perform non-local backbone movements ( AbRelax , Relax , Small ) . Proteins may have undergone local motions similar to Backrub and KIC moves to accommodate new mutations and amino acid interactions during evolution [24] , [35] , [36] , [41] . Such motions could have provided proteins with a mechanism to allow subtle , incremental changes to their structures without adversely affecting protein structure or protein function . While local motions may be a common mechanism for proteins to accommodate point mutations , larger structural adjustments may be necessary for dealing with insertions or deletions . In this study , we found that a moderate degree of backbone flexibility best recapitulated natural amino acid covariation , however , the magnitude of structural variation produced by this degree of backbone flexibility was less than the structural variation among naturally occurring protein families . This discrepancy is likely due to the assumption in the design method that the protein remains a fixed length . This is not true in naturally occurring sequences; in fact , all 40 domains in our benchmark include loop regions that have varying lengths . Mutations that change the length of a flexible loop could allow for secondary structure elements to re-orient themselves and slightly alter the tertiary structure . The accumulation of mutations in loop regions can produce significant structural diversity that cannot be modeled using a protein design method that keeps the number of amino acids in a protein constant . Future protein design methods , particularly those involving loop regions such as protein-protein interaction design or enzyme specificity design , could potentially benefit from incorporating moves that both change the conformation and length of the protein backbone . In addition to observing significant similarity between the sets of natural and designed highly covarying amino acid pairs , we observed a high correlation in the amino acid propensities of these covarying pairs and showed that the structural mechanisms underlying covariation are similar for both natural and designed sequences . Differences between natural and designed covarying pairs highlight areas for improvement in the energy function used for protein design . For instance , cation-pi interactions , which are not explicitly accounted for in the energy function used in this study , have high propensities among naturally covarying pairs but not in designed covarying pairs . Similarly , polar amino acid pairs are more frequent in the cores of designed proteins than in naturally occurring proteins . Interestingly , we found differences in the pair propensities between nature-specific pairs and pairs that highly covary in both natural and design sequences . We also observed that nature-specific pairs tend to be more distant in three-dimensional structure . These results have implications for the field of contact prediction , as combining amino acid covariation with amino acid pair propensity information could improve the prediction of three-dimensional contacts in protein structures compared to using amino acid covariation alone . Improving methods of contact prediction would increase the accuracy of recent protein structure prediction algorithms that use amino acid covariation [29] . Unlike nature-specific pairs , design-specific pairs have amino acid propensities that are highly correlated with the amino acid propensities of pairs that covary in both natural and designed sequences . These design-specific pairs represent candidate positions for engineering amino acid interactions that have not been sampled by natural protein evolution . A practical application of this is the re-wiring of protein interaction specificity to design orthogonal protein-protein interactions for use in synthetic biology . Natural intermolecular covariation has previously been exploited to alter specificity in two component signaling systems [42] . Future work could exploit designed intermolecular covariation to re-engineer protein interactions with novel specificities that are orthogonal from naturally occurring protein-protein interactions [43] and therefore useful for synthetic applications .
The protein domains used in this study were selected from the Pfam database [44] based on the following criteria: 1 ) at least one crystal structure of a protein containing the domain was available from the Protein Data Bank ( PDB ) [45] , 2 ) at least 500 sequences of proteins from the domain were available from Pfam and 3 ) the domain was equal to or less than 150 amino acids in length . We selected a total of 40 domains that represented a diverse set of protein folds ( Table 1 ) . The seed alignment and the full alignment for each domain were obtained from Pfam . In order to remove highly divergent sequences with uncommon insertions or deletions , we first removed sequences from the seed alignment if they had either of the following: 1 ) a gap in a position where 90% of the sequences in the seed alignment did not have a gap or 2 ) an amino acid in a position where 90% of the sequences in the seed alignment had a gap . Next , we aligned each sequence in the full alignment to the seed alignment using MUSCLE [46] and we discarded any sequences that resulted in the creation of gaps that were not in the seed alignment . This resulted in an alignment without sequences containing uncommon insertions or deletions . Finally , we used CD-HIT [47] to filter the sequence alignments by removing sequences with 80% redundancy or greater . For each of the 40 protein domains , the highest resolution crystal structure of the domain was obtained from the PDB . This structure was used as a template for all the design simulations . The design method used in this study consisted of two steps: 1 ) the generation of a conformational ensemble and 2 ) the design of sequences onto each structure in the ensemble using RosettaDesign . For each of the 40 domains , 500 structures were generated for the conformational ensemble and 500 sequences were designed , one for each structure in the ensemble . Descriptions of each protocol used for generating conformational ensembles and for designing sequences are provided in Text S1 along with the corresponding Rosetta command lines . Amino acid covariation was quantified using a mutual information based metric called Zpx [28] . First , the Shannon entropy is calculated at each position i as follows:where Px is the frequency of amino acid x at position i . The joint entropy is calculated between all pairs of positions as follows:where Px , y is the frequency of amino acid x and y and positions i and j , respectively . The mutual information ( MI ) between each pair of columns in a multiple sequence alignment , i and j , was calculated as the difference between the individual entropies and the joint entropy:Next , the background mutual information due to random noise and shared ancestry is subtracted to obtain the product corrected mutual information ( MIp ) [27]:where is the mean MI of position i with all other positions and is the overall mean . This value is converted to two Z-scores , one for each column , which are multiplied together:The final score , called Zpx , is the square root of the absolute value of . If is negative , then Zpx is multiplied by −1 . This normalization of MIp was demonstrated to reduce the sensitivity to misaligned regions in multiple sequence alignments , which otherwise result in artificially high mutual information scores [28] . Calculation of Zpx was implemented in Python . Direct coupling analysis ( DCA ) was calculated using Matlab code provided by its authors [37] . To compare amino acid covariation between natural and designed multiple sequence alignments , Zpx was first computed for all pairs of ungapped positions in each alignment . The mean Zpx for each alignment was calculated and residue pairs with values greater than two standard deviations above the mean Zpx were considered to be covarying residue pairs . The covariation similarity between the natural and designed covarying amino acid pairs was calculated as the percent of overlap , 2C/ ( A+B ) , where A and B are the total numbers of natural and designed covarying pairs , respectively , and C is the number of pairs that covary in both natural and designed sequences . The same approach was used to calculate covariation similarity using DCA . Sequence recovery was calculated as the mean percent identity of the designed sequences to the sequence of the crystal structure used as input for the design protocol . Sequence entropy was calculated for each position as defined above . Sequence profile similarity was calculated as the mean prof_sim score [48] between each position in the natural and designed alignments . Briefly , prof_sim is the product of two scores: 1 ) the estimated probability that two amino acid frequency distributions represent the same source distribution and 2 ) the a prior probability of the source distribution . Using this metric , positions in designed sequences receive high prof_sim scores if both 1 ) their amino acid distribution is similar to the amino acid distribution at the corresponding position in the natural alignment and 2 ) their amino acid distribution is different than the background amino acid distribution . Calculation of sequence recovery , entropy and profile similarity was implemented in Python . Structural variation was calculated as the mean pair-wise RMSD between 10 randomly selected structures in each conformational ensemble . Natural structural variation was computed for all domains with at least 10 crystal structures in the PDB . The following 20 domains were used to compute natural structural variation: PF00013 , PF00018 , PF00041 , PF00072 , PF00076 , PF00085 , PF00111 , PF00168 , PF00169 , PF00179 , PF00254 , PF00355 , PF00439 , PF00550 , PF00581 , PF00582 , PF00595 , PF01833 , PF07679 , PF07686 . Structural alignments and RMSD calculations were performed using PyMol [49] . Amino acid pair propensities ( PP ) were calculated as the ratio between observed pair frequencies and the expected individual amino acid frequencies:To compare amino acid pair propensities between two sets of covarying pairs , we computed the Z-score for each pair amino acid pair x , y . The Pearson correlation coefficient r between the two sets of Z-scores was then calculated using R [50] . Cysteines were excluded from this analysis because they rarely appear in the designed sequences . To classify the mechanisms of covariation for a pair of positions , we first computed a correlation coefficient for each amino acid pair x , y [32] . We then calculated a score for all possible amino acid pair transitions ( PT ) between one pair x , y and another pair a , b as follows:This pair transition score quantifies the significance of the transition between the amino acid pair x , y and the pair a , b . The most significant transitions are defined as those that highly favor pairs x , y and a , b but highly disfavor pairs x , b and a , y . For each pair of positions , ten pair transitions with the greatest scores were assigned one of eight classes in the following order: charge , cation-pi , pi-pi , size , hydrogen bonding , other ( hydrophobic ) , other ( hydrophilic ) and other ( mixed ) . Charge transitions involve a pair with opposite charges that either swap sign or become uncharged . A charge transition is also assigned to pair transitions that avoid like charges , for example , if x and b ( or y and a ) are like charges . Cation-pi transitions involve one pair with a potential cation-pi interaction but no cation-pi interaction in the other pair . Similarly , pi-pi transitions involve one pair with a potential pi-pi interaction but no pi-pi interaction in the other pair . Size transitions involve a decrease in the size of one amino acid by at least 18 Å3 ( the volume of a methyl group ) and an increase in the size of the other amino acid by at least 18 Å3 . Hydrogen bonding transitions involve a potential hydrogen bonding interaction ( hydrogen bond acceptor and donor ) in one pair but not in the other pair . The three other classes are used to assign pair transitions that do not fit any of the above criteria . Other ( hydrophobic ) transitions are those where both pairs contain only hydrophobic amino acids , other ( hydrophilic ) transitions are those where both pairs contain only hydrophilic amino acids , and other ( mixed ) transitions are those with both hydrophobic and hydrophilic amino acids . Similarity between natural and designed was quantified using the percent overlap ( defined above ) for each covariation mechanism . Amino acid burial was defined for each position based on the number of Cβ atoms within 8 Å of the Cβ atom of the given position as follows: exposed 0–8 , intermediate 9–14 and buried >14 . For the covariation mechanism analysis in Figure S4 , we defined pairs of positions that were buried/buried or buried/intermediate as buried pairs , exposed/buried or intermediate/intermediate as intermediate pairs , and exposed/intermediate or exposed/exposed as exposed pairs . For domains with known protein–ligand or protein–protein interface information , we defined all positions with a heavy-atom within 6 Å of any heavy-atom on the binding partner as an interface position . The domains with interface information were PF00013 , PF00439 , PF00498 , PF00691 , PF00072 , PF00018 , PF00076 , PF00249 , PF00327 , PF01035 , PF00169 , PF00550 and PF00595 . For domains with known active sites , we defined all positions with a heavy-atom within 6 Å of any heavy-atom on a catalytic residue as an active site position . The domains with active site information were PF00085 , PF00111 , PF00355 , PF00708 , PF00581 , and PF01451 . | Proteins generally fold into specific three-dimensional structures to perform their cellular functions , and the presence of misfolded proteins is often deleterious for cellular and organismal fitness . For these reasons , maintenance of protein structure is thought to be one of the major fitness pressures acting on proteins . Consequently , the sequences of today's naturally occurring proteins contain signatures reflecting the constraints imposed by protein structure . Here we test the ability of computational protein design methods to recapitulate and explain these signatures . We focus on the physical basis of evolutionary pressures that act on interactions between amino acids in folded proteins , which are critical in determining protein structure and function . Such pressures can be observed from the appearance of amino acid covariation , where the amino acids at certain positions in protein sequences are correlated with each other . We find similar patterns of amino acid covariation in natural sequences and sequences optimized for their structures using computational protein design , demonstrating the importance of structural constraints in protein molecular evolution and providing insights into the structural mechanisms leading to covariation . In addition , these results characterize the ability of computational methods to model the precise details of correlated amino acid changes , which is critical for engineering new proteins with useful functions beyond those seen in nature . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [] | 2013 | Computational Protein Design Quantifies Structural Constraints on Amino Acid Covariation |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.